Content Delivery Network (CDN) Service Specific Terms

1. Definitions

Number of 5xx response errors in five minutes: Number of 5xx errors returned by the domain name due to CDN system problems in five minutes.

Number of failed requests due to node unavailability in five minutes: If a node is unavailable, calculate the number of requests per five minutes based on the statistics of the first seven days, and then convert it to the number of failed requests due to node unavailability in five minutes.

Error in five minutes = (Number of 5xx response error in five minutes + Number of failed requests due to node unavailability in five minutes) / Total number of requests in five minutes of the domain name.

"Service Unavailability" means the CDN system log shows that the error rate of CDN was greater than 0.05% within five (5) minutes because of Huawei Cloud.

2. Disclaimers

If you will have a sharp increase in CDN bandwidth usage (greater than or equal to 30% of the billed bandwidth for last month), you must submit a service ticket to inform us at least three (3) business days in advance. Otherwise, HUAWEI CLOUD will not be responsible for service unavailability resulting from this.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of CDN per Service Cycle not less than 99.9%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the below service credit:

Service Availability	Service Credit
99%≤Service Availability Rate <99.9%	10%
Service Availability Rate <99%	25%

Last Updated: January 15, 2024

Anti-DDoS Service (AAD) Service Specific Terms

1. Definition

Service Unavailability is the status that an AAD system was inaccessible and incapable of forwarding traffic for five (5) or more consecutive minutes. Duration of the inaccessibility and incapability that lasts shorter than five (5) minutes shall not be included.

2. Disclaimers

This SLA does not apply to any performance or availability issues: 

1) Due to security threats or fraud or illegal acts caused by you or your end users to the Services;

2) Issues in the back-end origin server, such as fully occupied bandwidth, exposed IP address, data center faults, and/or link jitter;

3) Black holes caused by attack traffic exceeding your purchased AAD capability.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of AAD per Service Cycle not less than 99.9%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the below service credit:

Service Availability Rate	Service Credit
99.8%≤Service Availability Rate <99.9%	10% of Monthly Service Fee
99.5%≤Service Availability Rate <99.8%	20% of Monthly Service Fee
Service Availability Rate <99.5%	50% of Monthly Service Fee

Last Updated: January 15, 2024

Elastic Cloud Server (ECS) Service Specific Terms

1. Definition

Service Cycle: A service cycle is a calendar month.

Total Time of Service Cycle per ECS is the total number of minutes in a Service Cycle for an ECS. Total Time of Service Cycle = Total number of days in a Service Cycle x 24 hours x 60 minutes.

ECS Unavailability is when the system logs of Elastic Cloud Servers (ECSs) (including FlexusX instances and ECSs on DeHs) show that ECS was inaccessible for one (1) or more consecutive minutes because of Huawei Cloud.

Unavailable Time per ECS is the total number of minutes that an ECS was unavailable in a Service Cycle.

Multi-AZ ECS Unavailability in One Region is the unavailability of ECSs in other availability zones (AZs) when the ECSs in one AZ in the same region are available.

Multi-AZ ECS Unavailability Time in One Region is the total number of minutes that the ECSs in other AZs experienced Multi-AZ ECS Unavailability in One Region within a Service Cycle.

Monthly Expenditure is your total expenditure for an ECS within a Service Cycle. If you pay for several months at a time, the Monthly Expenditure is calculated by dividing the total amount billed by the number of months the service was used.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Data stored on local disks being lost due to damage to those disks or becoming unavailable due to an instance breakdown;

2) If your service does not have high availability (HA) capabilities and does not accept server replacement,BMS or ECS with local disks instance becomes unavailable during hardware repair;

3) Service unavailability caused by your own operations, for example, stopping or restarting an ECS, or detaching an EVS disk from an ECS using the console or APIs;

4) Services made unavailable due to unsupported software or drivers, third-party software or drivers that are not directly managed by Huawei Cloud, configuration errors, or hacker attacks;

5) Data, passwords, or other information getting lost or leaked due to improper maintenance or handling of confidential information;

6) Spot ECSs released because the bid price was lower than the market price or ECSs suspended or released due to payment in arrears;

7) Unavailability caused by a force majeure;

8) Unavailability caused by your operation errors or operations authorized by you;

9) Network faults, device faults, or configuration adjustment of non-Huawei devices;

10) Short-term server interruption or network unavailability caused by live migration (which is performed in advance to prevent server breakdown caused by hardware faults);

11) Unavailability caused by your failure to respond to Huawei Cloud notices on routine maintenance activities such as repair, maintenance, upgrades, and optimization of service platforms or related equipment, systems, and software.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

The ECS Service Availability Rate can be calculated by referring to the following formula:

1) Single ECS service availability (ECS availability within a given AZ and region)

Service Availability Rate per Service Cycle = (Total Time of a Service Cycle per ECS – Unavailable Time per ECS)/Total Time of Service Cycle per ECS x 100%.

2) Multi-AZ ECS service availability in one region (ECS availability in multiple AZs in a region)

Service Availability Rate per Service Cycle = (Total Time of a Service Cycle per ECS – Multi-AZ ECS Unavailability Time in One Region)/Total Time of Service Cycle per ECS x 100%.

Service Availability Rate Commitment

1) Huawei Cloud will use all commercially and technically reasonable efforts to ensure that the Service Availability Rate of a Single ECS (ECS availability within a given AZ and region) per Service Cycle is not lower than 99.975% and that the Multi-AZ ECS Availability Rate in One Region per Service Cycle is not lower than 99.995%. You are advised to deploy ECSs in at least two AZs in one region (if a given region has only one AZ, then deploy your ECSs in at least two regions) for a higher service availability rate. If you choose to deploy your ECSs in a single AZ in a given region, Huawei Cloud shall only promise a single ECS service availability (ECS availability within a given AZ and region) per Service Cycle.

2) ECS provides different deployment modes for your applications. You can deploy one or more applications on a single ECS, or you can deploy applications in active/standby or cluster mode on multiple ECSs. You have understood and agreed that different deployment modes deliver different levels of reliability. Choose an appropriate one to meet your requirements for service reliability. To prevent single points of failure (SPOFs) from affecting core services, you are advised not to deploy HA applications on a single ECS.

4. Compensation

Compensation Standard

If we fail to meet the commitment described here for the ECS Service Availability Rate, Huawei Cloud will compensate you with the service credits described below (only for purchasing ECSs). The total amount of the service credits shall not exceed the Monthly Service Fee minus any coupons that have been applied.

Single ECS availability (ECS availability within a given AZ and region)

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.975%	10% of ECS expenditures for that month
95% ≤ Service Availability Rate < 99%	30% of ECS expenditures for that month
Service Availability Rate < 95%	100% of ECS expenditures for that month

Multi-AZ ECS service availability in one region

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.995%	10% of ECS expenditures for that month
95% ≤ Service Availability Rate < 99%	30% of ECS expenditures for that month
Service Availability Rate < 95%	100% of ECS expenditures for that month

Compensation Application Time Limit

You can claim compensation for an ECS that did not meet Service Availability requirements for a given month after the fifth (5) business day of the following month. The claim for compensation must be made within two (2) months after the end of the month in which Service Availability was not met. Claims for compensation beyond this time limit will not be accepted.

 

Last Updated: August 16, 2024

 

You can see what is updated in Elastic Cloud Server (ECS) Service Level Agreement History Version.

Elastic Load Balance (ELB) Service Specific Terms

1. Definition

Service Unavailability means the ELB system log shows that ELB was inaccessible for one (1) or more consecutive minutes or ELB failed to distribute the traffic when at least one backend server is healthy.

Monthly Service Fee means service fees paid by you for ELB in a Service Cycle. For example, if you pay the service fee of several months at a time, the Monthly Service Fee is calculated as the total service fees paid divided by the total number of months.

2. Disclaimers

This SLA does not apply to any performance and availability issues:

1) caused by force majeure;

2) that arise during the period when your services are suspended or terminated according to the laws and regulations at the request of supervision authorities or the customer agreement;

3) caused by you or any third party;

4) caused by daily service maintenance;

5) that arise due to your failure to comply with the operation guidance of the services;

6) caused by a serious problem of the network operator;

7) caused by planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

8) caused by hacker attacks on your applications, including but not limited to DDoS attacks, SYN flood attacks, and virus attacks;

9) caused by a faulty backend server, or the absence of backend servers;

10) caused by the overloaded traffic that your current load balancer specification cannot handle;

11) caused by your security policy that denies health check to monitor the health of backend servers;

12) caused by loss or disclosure of data, including passwords, due to your improper maintenance;

13) caused by your negligence or allowed operations.

3. Service Availability

Formula for calculating Service Availability in each Service Cycle:

Service Availability in each Service Cycle = (Total Time of Service Cycle – Service Unavailability Duration) / Total Time of Service Cycle x 100%.

Service Availability commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability of ELB per Service Cycle be as follows:

Load Balancer Type	Service Availability (Single AZ)	Service Availability (Multiple AZs)
Dedicated load balancers	99.9%	99.99%
Shared load balancers (Guaranteed performance enabled)	—	99.95%
Shared load balancers	No SLA commitment

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the below service credit:

Load Balancer Type	Service Availability (Single AZ)	Service Availability (Multiple AZs)	Service Credit
Dedicated load balancers	99% ≤ Service Availability < 99.9%	99% ≤ Service Availability < 99.99%	10% of Monthly Service Fee
	Service Availability < 99%	Service Availability < 99%	30% of Monthly Service Fee
Shared load balancers (Guaranteed performance enabled)	—	99% ≤ Service Availability < 99.95%	10% of Monthly Service Fee
		Service Availability < 99%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Elastic Volume Service (EVS) Service Specific Terms

1. Definition

Service Unavailability means the Elastic Volume Service (EVS) system log shows that EVS was inaccessible for one (1) or more consecutive minutes because of Huawei Cloud.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of EVS per Service Cycle not less than 99.95%.

3. Service Credit

If we fail to meet the commitment above on EVS Service Availability Rate, Huawei Cloud will provide you with the below service credit: 

Service Availability Rate	Service Credit
99%≤Service Availability Rate <99.95%	10% of Monthly Service Fee
95%≤Service Availability Rate <99%	30% of Monthly Service Fee
Service Availability Rate <95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Object Storage Service (OBS) Service Specific Terms

1. Definitions 

Error Rate Per Five Minutes: For different regions and storage classes, the Error Rate Per Five Minutes is calculated as follows:

Error Rate Per Five Minutes = Total number of Failed Requests per five minutes/Total number of Valid Requests per five minutes * 100%.

Failed Requests: Refer to requests with 5xx error codes returned by OBS, except the following:

1) Requests whose client IP addresses and accounts are restricted by OBS due to hacker attacks on your resources;

2) Requests (with an HTTP status code 429 or 503) that exceed the specified quota due to improper access, and are therefore restricted by OBS;

3) Failed requests generated when the GetService API is called to obtain the bucket list;

4) Failed requests caused by the backend asynchronous processing after lifecycle management and cross-region replication are configured;

5) Slow requests in the following scenarios:

a)  Callback timeout requests;

b)  Timeout requests of back-to-source by mirroring;

6) Error requests due to appropriate upgrade, change, or suspension initiated by OBS.

Valid Requests: Refer to requests received by the OBS server, except the following:

1) Requests failed (with HTTP status code 403) to pass OBS identity verification and authentication;

2) Requests failed (with HTTP status code 403) due to account arrears;

3) Requests calling the GetService API to obtain the bucket list;

4) Requests initiated by the backend asynchronous processing after lifecycle management and cross-region replication are configured;

5) Requests crafted by hacker attacks on your applications.

2. Disclaimers

This SLA does not apply to any performance or availability issues: Edge access failures due to network faults (such as edge node network disconnections and carrier line's delay jitter or exceptions), hardware faults (for example, device's power supply faults), data center faults, and others.

3. Service Availability

3.1 Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate equals to 1 minus the average error rate and then multiplied by 100%. The average error rate is calculated by dividing the sum of Error Rate Per Five Minutes in a Service Cycle with the total number of five-minute cycles in the Service Cycle. The formula is as follows:

Service Availability Rate in each Service Cycle = (1 – ∑Error Rate Per Five Minutes in the Service Cycle/Total number of five-minute cycles in the Service Cycle) * 100%.

(Note: Total number of five-minute cycles in the Service Cycle = Total number of days in the Service Cycle * 24 (hours) * 12).

3.2 Service Availability Rate commitment

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of Standard storage in a single availability zone (AZ) per Service Cycle no less than 99.9%, Standard storage in three AZs per Service Cycle no less than 99.995%, Infrequent Access storage in a single AZ per Service Cycle no less than 99%, Infrequent Access storage in three AZs per Service Cycle no less than 99.5%, and Archive storage per Service Cycle no less than 99%.

4. Service Credit

If we fail to meet the above commitment on OBS Service Availability Rate, Huawei Cloud will provide you with the below Service Credit according to different storage classes in different AZ deployment:

Standard stored in a single AZ: 

Service Availability Rate	Service Credit
99．9%≤Service Availability Rate <99.99%	10% of Monthly Service Fee
95%≤Service Availability Rate <99.9%	30% of Monthly Service Fee
Service Availability Rate <95%	50% of Monthly Service Fee

Standard stored in three AZs: 

Service Availability Rate	Service Credit
99．9%≤Service Availability Rate <99.995%	10% of Monthly Service Fee
95%≤Service Availability Rate <99.9%	30% of Monthly Service Fee
Service Availability Rate <95%	50% of Monthly Service Fee

Infrequent Access storage in a single AZ and Archive storage : 

Service Availability Rate	Service Credit
95%≤Service Availability Rate <99%	30% of Monthly Service Fee
Service Availability Rate <95%	50% of Monthly Service Fee

Infrequent Access storage in three AZs:

Service Availability Rate	Service Credit
99%≤Service Availability Rate <99.5%	10% of Monthly Service Fee
95%≤Service Availability Rate <99%	30% of Monthly Service Fee
Service Availability Rate <95%	50% of Monthly Service Fee

Last Updated: January 15, 2024

RDS for MySQL, RDS for PostgreSQL, RDS for SQL Server, and RDS for MariaDB Service Specific Terms

1. Definition

Service Unavailability: The RDS system log shows that a DB instance was inaccessible for one or more consecutive minutes because of Huawei Cloud.

2. Disclaimers

Service unavailability caused by the following situations is not be counted as part of the service unavailability duration:

1) User operations, such as rebooting, deleting, or restoring DB instances through a console or APIs;

2) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, and fault simulations;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Service interruptions caused by insufficient vCPUs, memory, storage, or the maximum number of connections being reached for a DB instance or a database proxy due to heavy customer workloads;

5) Hacker attacks on user applications;

6) Consequences of data (including passwords) leakage or loss due to improper maintenance of the customer;

7) Operations authorized by the customer or any mis-judgment in operations;

8) Customer failure to comply with the product usage documentation or recommendations provided by Huawei Cloud;

9) Redo or recovery log replay;

10) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

11) Data losses resulting from a local disk failure;

12) Force majeure factors;

13) Faults resulting from single-AZ deployment instead of multi-AZ deployment;

14) Database engine basic software that causes databases to break down repeatedly or to be inoperable;

15) Faults caused by open-source community bugs or other non-Huawei Cloud issues.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to ensure the Service Availability Rate of RDS primary/standby instances is no less than 99.95% per Service Cycle.

Note: The availability (SLA) cannot be guaranteed if the instance (including read replicas) is deployed on a single server.

4. Service Credit

If we fail to meet this RDS Service Availability commitment, Huawei Cloud will provide you with the service credits listed below:

Primary/Standby instances

Service Availability	Service Credit
99% ≤ Service Availability < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability < 99%	25% of Monthly Service Fee
Service Availability < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Data Lake Insight (DLI) Service Specific Terms

1. Definition

Service Unavailable means the DLI system log shows that DLI was inaccessible for more than five consecutive minutes due to a fault with Huawei Cloud.

2. Disclaimers

This Statement does not apply to any performance or availability issues caused by the following:

1) Customer operations, for example, deleting, stopping, or restarting a cluster through the console or APIs, or deleting a job, connection, or driver;

2) Planned system maintenance, including but not limited to cutover, maintenance, upgrade, patching, and fault simulations, of which Huawei Cloud has notified the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Hacker attacks on user applications;

5) Consequences of data (including passwords) leakage or loss due to improper maintenance of the customer;

6) Operations authorized by the customer;

7) Customer's failure to comply with the product usage documentation or recommendations provided by Huawei Cloud;

8) Queue interruption caused by insufficient CPU and storage resources due to heavy customer service workload;

9) Service interruption caused by an unavailable OBS pool selected by the customer;

10) service interruption caused by other customer factors. For example, a job developed by the customer fails to be executed.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use all commercially and technically reasonable efforts to ensure that the Service Availability Rate of a DLI queue (in multiple AZs) per Service Cycle is not lower than 99.95% and that of a DLI queue (in a single AZ) is not lower than 99.90%.

4. Compensation

If we fail to meet the commitment above on DLI Service Availability Rate, Huawei Cloud will compensate you as showed in the following table:

DLI Queue (Single AZ)
Service Availability Rate	Coupon Amount
99.00% ≤ Service Availability Rate < 99.90%	10% of Monthly Expenditure
95.00% ≤ Service Availability Rate < 99.00%	30% of Monthly Expenditure
Service Availability Rate < 95.00%	50% of Monthly Expenditure
DLI Queue (Multiple AZs)
Service Availability Rate	Coupon Amount
99.90% ≤ Service Availability Rate < 99.95%	10% of Monthly Expenditure
95.00% ≤ Service Availability Rate < 99.90%	30% of Monthly Expenditure
Service Availability Rate < 95.00%	50% of Monthly Expenditure

Last Updated: January 15, 2024

Cloud Stream Service (CS) Service Specific Terms

1. Definition

Service Unavailability means the CS system log shows that a CS was inaccessible for five (5) or more consecutive minutes because of Huawei Cloud. Duration of the inaccessibility and inability that lasts for shorter than five (5) minutes shall not be calculated as Service Unavailability.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of CS per Service Cycle no less than 99.95%.

3.Service Credit

If we fail to meet the commitment above on CS Service Availability Rate, Huawei Cloud will provide you with the below service credit: 

Service Availability Rate	Service Credit
99%≤Service Availability Rate <99.95%	10% of Monthly Service Fee
95%≤Service Availability Rate <99%	30% of Monthly Service Fee
Service Availability Rate <95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Elastic IP (EIP) Service Specific Terms

1. Definition

Service Unavailability means the Elastic IP (EIP) system log showing that all data packets in the outbound direction are discarded at the Huawei Cloud egress gateway for one (1) or more consecutive minutes due to the fault of Huawei Cloud.

Monthly Expenditure is your total expenditures for a single EIP within a Service Cycle (a calendar month). If you pay for several Service Cycles at a time, the Monthly Expenditure is calculated by dividing the total amount billed by the number of months the EIP was used.

2. Disclaimers

Service Unavailability does not include unavailability resulting from any of the following:

1) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

2) Non-Huawei Cloud network failures, device faults, or configuration changes, including but not limited to service unavailability caused by network carrier line faults;

3) Application attacks including but not limited to DDoS attacks, SYN flood attacks, and virus attacks, or software and hardware compatibility issues;

4) Your services exceeding the supported resource specifications;

5) Faulty backend servers;

6) Loss or disclosure of data, including passwords, due to your improper maintenance;

7) Operations authorized by you or any mis-judgment in operations;

8) Your failure to follow to the product usage documentation or recommendations provided by Huawei Cloud;

9) Force majeure or other accidents that are unforeseeable, insurmountable, and unavoidable and have significant impact on one party or both parties, including but not limited to natural disasters (such as floods, earthquakes, and plague epidemics) and social events (such as wars, unrest, government actions, interruption of telecom backbone lines, hackers, network congestion, technical change of telecommunications departments, and government policies).

3. Service Availability

Service Availability Rate in each Service Cycle is represented by the following formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to ensure the Service Availability Rate of dynamic BGP EIPs, static BGP EIPs, and EIPs of China Telecom, China Unicom, or China Mobile per Service Cycle are, respectively, no less than 99.95%, 99%, and 99%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the service credits described below:

Dynamic BGP EIPs (including Premium BGP, International BGP, Local BGP, etc.):

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Static BGP EIPs:

Service Availability Rate	Service Credit
98% ≤ Service Availability Rate < 99%	10% of Monthly Service Fee
Service Availability Rate < 98%	25% of Monthly Service Fee

EIPs of China Telecom, China Unicom, or China Mobile:

Service Availability Rate	Service Credit
98% ≤ Service Availability Rate <99%	10% of Monthly Service Fee
Service Availability Rate < 98%	25% of Monthly Service Fee

Last Updated: July 31, 2024

Direct Connect (DC) Service Specific Terms

Self-Service Connections

1. Definitions

Direct Connect Location: a Huawei Cloud data center that can access Huawei Cloud.

Standard Scenario: multiple Direct Connect connections accessing multiple locations.

Service Unavailability: All attempts to connect to a specified Direct Connect connection within one minute have failed. Service Unavailability refers to the total minutes in a Service Cycle in which all attempts to connect to a specified Direct Connect connection failed.

2. Disclaimers

Service Unavailability does not include unavailability resulting from any of the following:

1) Service fluctuations due to planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

2) Migration of your resources due to Huawei Cloud equipment room migration or device upgrade;

3) One or more connections being connected to a single Direct Connect location that has failed;

4) Service interruptions caused by network failures, device faults, or configuration changes in non-Huawei Cloud data centers, including but not limited to service unavailability caused by network carrier line faults, or your device or network faults;

5) Direct Connect being unavailable due to malfunctions of other services;

6) Loss or disclosure of data, including passwords, due to your improper maintenance;

7) Operations authorized by you or any mis-judgment in operations;

8) Your failure to follow the product usage documentation or recommendations provided by Huawei Cloud;

9) Your services exceeding the supported resource specifications;

10) Application attacks including but not limited to DDoS attacks, SYN flood attacks, and virus attacks, or software and hardware compatibility issues;

11) Force majeure or other accidents that are unforeseeable, insurmountable, and unavoidable and have significant impact on one party or both parties, including but not limited to natural disasters (such as floods, earthquakes, and plague epidemics) and social events (such as wars, unrest, government actions, interruption of telecom backbone lines, hackers, network congestion, technical change of telecommunications departments, and government policies).

Note: This SLA only ensures the service availability of Direct Connect connections terminated at different Direct Connect locations. If Huawei Cloud needs to migrate your resources due to equipment room migration or device upgrade, Huawei Cloud will notify you in advance, and you are obligated to cooperate with Huawei Cloud.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Direct Connect availability refers to the availability of the network connections between Direct Connect locations and Huawei Cloud VPCs. To ensure the availability of self-built lines, your need to contact your carriers.

Huawei Cloud will make all commercially viable and technically feasible efforts to ensure that the Service Availability Rate of Direct Connect per Service Cycle is not less than 99.95%, but will only do so in multiple-location access scenarios.

4. Service Credit

If we fail to meet these commitments, Huawei Cloud will provide you with the service credits described below:

Service Availability Rate	Service Credit
99.5% ≤ Service Availability Rate < 99.95%	5% of Monthly Service Fee
99% ≤ Service Availability Rate < 99.5%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

------------------------------------------------------------------------------------

Full-Service Connections

1. Definitions

Direct Connect Location: a Huawei Cloud data center that can access Huawei Cloud.

Full-Service Connection: Huawei Cloud and third-party network carriers jointly provide end users with convenient cloud services. Huawei Cloud provides Direct Connect and some network proxy services. Third-party network carriers (hereafter referred to as "carriers") provide physical line-related services.

Standard Scenario: multiple Direct Connect connections accessing multiple locations

Service Unavailability: All attempts to connect to a specified Direct Connect connection within one minute have failed. Service Unavailability refers to the total minutes in a Service Cycle in which all attempts to connect to a specified Direct Connect connection failed.

2. Disclaimers

Service Unavailability does not include unavailability resulting from any of the following:

1) Service fluctuations due to planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

2) Migration of your resources due to Huawei Cloud equipment room migration or device upgrade;

3) One or more connections being connected to a single Direct Connect location that has failed;

4) Service interruptions caused by network failures, device faults, or configuration changes in non-Huawei Cloud data centers, including but not limited to service unavailability caused by network carrier line faults, or your device or network faults;

5) Direct Connect being unavailable due to malfunctions of other services;

6) Loss or disclosure of data, including passwords, due to your improper maintenance;

7) Operations authorized by you or any mis-judgment in operations;

8) Your failure to follow the product usage documentation or recommendations provided by Huawei Cloud;

9) Your services exceeding the supported resource specifications;

10) Application attacks including but not limited to DDoS attacks, SYN flood attacks, and virus attacks, or software and hardware compatibility issues;

11) Force majeure or other accidents that are unforeseeable, insurmountable, and unavoidable and have significant impact on one party or both parties, including but not limited to natural disasters (such as floods, earthquakes, and plague epidemics) and social events (such as wars, unrest, government actions, interruption of telecom backbone lines, hackers, network congestion, technical change of telecommunications departments, and government policies).

Note: This SLA only ensures the service availability of Direct Connect connections terminated at different Direct Connect locations. If Huawei Cloud needs to migrate your resources due to equipment room migration or device upgrade, Huawei Cloud will notify you in advance, and you are obligated to cooperate with Huawei Cloud.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Direct Connect availability refers to the availability of the network connections between Direct Connect locations and Huawei Cloud VPCs. To ensure the availability of self-built lines, your need to contact your carriers.

4. Service Credit

If we fail to meet these commitments, Huawei Cloud will provide you with the service credits described below:

Service Availability Rate	Service Credit
99.5% ≤ Service Availability Rate < 99.95%	5% of Monthly Service Fee
99% ≤ Service Availability Rate < 99.5%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Document Database Service (DDS) Service Specific Terms

1. Definition

"Service Unavailability" means the DDS system log showing that DDS was inaccessible for five (5) or more consecutive minutes due to the fault of Huawei Cloud.

2. Disclaimers

For DDS, Service Unavailability in the following conditions cannot be considered Service Unavailable Duration, nor can the following conditions described in the definition of Service Unavailability:

1) Unavailability caused by the customer's failure to connect to DDS instances in a high availability mode;

2) Network faults, device faults, or configuration adjustment of non-Huawei devices;

3) Issues caused by improper service use that deviates from the instructions provided in documents such as Best Practices and FAQs;

4) Faults that occur due to the use of single node instances, instead of HA instances such as replica set or cluster instances;

5) Faults that occur due to single-AZ deployment, instead of multi-AZ deployment;

6) Issues caused by third-party open-source software;

7) Customer operations, such as restarting, deleting, or restoring DB instances through a console or APIs;

8) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

9) Network faults, device faults, or configuration adjustment of non-Huawei devices;

10) Service interruption caused by insufficient CPU, memory, and storage resources due to customer's heavy service workload;

11) Hacker attacks on customer's applications;

12) Consequences of data (including passwords) leakage or loss due to improper maintenance of the customer;

13) Consequences caused by operations authorized by the customer or any misjudgment in operations;

14) Customer's failure to comply with the product usage documentation, development specifications, or recommendations provided by Huawei Cloud;

15) Faults caused by bugs in the open-source community or other non-Huawei Cloud issues;

16) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

17) Unavailability caused by operations authorized by the customer;

18) Other force majeure factors.

3. Service Availability

Service Availability Rate in each Service Cycle is represented by the following formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of DDS per Service Cycle be no less than 99.95%.

4.Service Credit

If we fail to meet the above commitment, we will provide you the service credit in the following table:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of the Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of the Monthly Service Fee
Service Availability Rate < 95%	100% of the Monthly Service Fee

Last Updated: January 15, 2024

GeminiDB Service Specific Terms

1. Description

GeminiDB is a distributed, multi-mode NoSQL service with decoupled compute and storage architecture developed by Huawei.

2. Definitions

"Service Unavailability" means the situation of GeminiDB system log showing that GeminiDB was inaccessible for five (5) or more consecutive minutes because of Huawei Cloud.

3. Disclaimers

For GeminiDB, the following situations will not be considered Service Unavailable Duration:

(1) Unavailability caused by the customer's failure to connect to GeminiDB instances in a high availability mode;

(2) Issues caused by improper service use that deviates from the instructions provided in documents such as Best Practices and FAQs;

(3) Faults that occur due to the use of single node instances, instead of HA instances such as replica set or cluster instances;

(4) Faults that occur due to single-AZ deployment, instead of multi-AZ deployment;

(5) Issues caused by third-party open-source software;

(6) Customer operations, such as restarting, deleting, or restoring DB instances through a console or APIs;

(7) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

(8) Network faults, device faults, or configuration adjustment of non-Huawei devices;

(9) Service interruption caused by insufficient CPU, memory, and storage resources due to customer's heavy service workload;

(10) Hacker attacks on customer's applications;

(11) Consequences of data (including passwords) leakage or loss due to improper maintenance of the customer;

(12) Consequences caused by operations authorized by the customer or any misjudgment in operations;

(13) Customer's failure to comply with the product usage documentation, development specifications, or recommendations provided by Huawei Cloud;

(14) Faults caused by bugs in the open-source community or other non-Huawei Cloud issues;

(15) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

(16) Unavailability caused by operations authorized by the customer;

(17) Unavailability caused by known problems or risks that have been reported to customers for rectification but still fail to be resolved;

(18) GeminiDB service availability is undermined due to acts, events, omissions, or incidents arising from events that are unforeseeable, inevitable, and beyond its reasonable control, including but not limited to natural disasters (earthquakes, storms, etc.), act of terrorism, war or warlike operations, civil unrest or riots, electricity outages, network or telecommunication faults, blockages, embargoes, fires, floods, explosions or malicious damage (including but not limited to hacker attacks), plant or equipment failure, or changes of any law, government order, rule, regulation, direction, or industry standard.

4. Service Availability

Service Availability Rate in each Service Cycle is represented by the following formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

5. Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of GeminiDB per Service Cycle be no less than 99.95%.

6. Service Credit

If we fail to meet the above commitment, we will provide you the service credit in the following table:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of the Monthly Service Fee
95%≤ Service Availability Rate < 99%	25% of the Monthly Service Fee
Service Availability Rate < 95%	100% of the Monthly Service Fee

Last Updated: September 12, 2024

You can see what is updated in GeminiDB Service Level Agreement History Version

ModelArts Service Specific Terms

1. Definitions

Service Unavailability: according to the log records in the ModelArts system, ModelArts service is inaccessible for 10 consecutive minutes due to Huawei Cloud and any unavailability period for less than 10 minutes is not included.

2. Service Availability

Formula for calculating Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration * 100%.

Commitment for Service Availability:

Huawei Cloud will ensure the service availability rate in each service cycle is at least 99.95% through technically and commercially reasonable efforts.

3. Restrictions

Except for the circumstances specified in the General Terms, with respect to ModelArts service, the following circumstances shall not be included in service unavailability time either, if the unavailability is caused by:

1) operations by clients (delete/suspend/restart or otherwise operate the resources through control console/API and other method, including work space, authorization, resource pool, instance in environment for development, data set, algorithm, training exercise, model, inference service);

2) system maintenance with prior notice from Huawei Cloud to clients, including but not limited to cutover, maintenance, upgrade, patch and simulated troubleshooting;

3) network or equipment failure or configuration adjustment not involving equipment in Huawei Cloud;

4) hacking on client applications;

5) loss or leakage of data, password, token due to improper maintenance or lack of confidentiality of client;

6) operations authorized by clients;

7) client failure in observing product documentation or usage advice of Huawei Cloud;

8) third-party open source software;

9) overload of client business, requests are flow controlled as requests exceed the specified quota;

10) client's own code issue or custom mirror issue;

11) third party dependencies (database and cache) introduced by client itself;

12) for client data that is stored in OBS bucket/SFS document system, the unavailability of OBS bucket/SFS document system of clients;

13) for clients who use self-managed SFS as the storage mounting in development environment, service exception caused by the unavailability of such SFS;

14) for clients who use EVS storage in development environment and back up data, data loss caused by the deletion of instance in development environment by clients;

15) the breakdown of one single node as the special resource pool created by clients is fewer than 2 nodes;

16) the breakdown of one single instance as one online service deployed by clients is fewer than 2 instances.

4. Service Credit

If ModelArts service availability fails to satisfy foregoing commitments, Huawei Cloud will compensate you in accordance with the following table:

Service Availability	Credit Amount
95.00% ≤ Service Availability < 99.95%	10% of monthly service fee
Service Availability < 95.00%	25% of monthly service fee

Last Updated: January 15, 2024

Cloud Backup and Recovery (CBR) Service Specific Terms

1. Definitions

Service Unavailability means the CBR system log shows that CBR was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of CBR per Service Cycle not less than 99.9%.

3. Service Compensation

If we fail to meet the above commitment, we will provide you the service credit in the following table:

Service   Availability Rate	Service   Credit
99% ≤ Service Availability Rate   < 99.9%	10% of Monthly Service Fee
95% ≤ Service Availability Rate   < 99%	25% of Monthly Service Fee
Service Availability Rate <   95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Data Warehouse Service GaussDB (DWS) Service Specific Terms

1. Definition

Service Unavailability means the GaussDB(DWS) system log shows that GaussDB(DWS) was inaccessible for more than five consecutive minutes due to a fault with Huawei Cloud.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) User operations, such as rebooting, deleting, or restoring DB instances through a console or APIs;

2) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) DWS service interruptions caused by an unavailable OBS pool selected by the customer;

5) Traffic limiting or fault that is caused by frequent access to management-plane interfaces but does not affect services;

6) Faults in test clusters (including POC clusters and other test clusters confirmed by the customer);

7) Service interruption caused by insufficient CPU and storage resources due to heavy customer service workload;

8) Unavailability caused by an AZ fault in a cluster where the customer does not enable 3-AZ HA;

9) Hacker attacks on user applications;

10) Consequences of data (including passwords) leakage or loss due to the fault of the customer;

11) Operations authorized by the customer;

12) Customer's failure to comply with the product usage documentation or recommendations provided by Huawei Cloud;

13) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

14) Other force majeure factors.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate Commitment:

Huawei Cloud will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of GaussDB(DWS) per Service Cycle is not less than 99.9%.

Note: Standalone deployment does not ensure availability.

4.Service Credit

If we fail to meet the commitment above on GaussDB(DWS) Service Availability Rate, Huawei Cloud will provide you with the following service credits:

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.90%	10% of Monthly Service Fee
Service Availability Rate < 99%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Cognitive Engagement Center (CEC) Service Specific Terms

1. Definition

Service Unavailability means inaccessible incoming and outgoing calls for five minutes or longer because of Huawei Cloud. The unavailability information is printed in customer service system (CSS) logs. The unavailability duration shorter than five minutes shall not be calculated.

2. Disclaimers

In addition to what are specified in the General Terms, this SLA does not apply to any performance or availability issues:

1) caused by your network faults, including changes on firewalls and network devices, and faults on network devices, agent computers, or internal office networks;

2) caused by your local voice gateway or SIP line failures when the phone number is provided by yourself;

3) caused by any leased line network failures if your local site communicates with Huawei cloud servers through a leased line which is purchased from a carrier.

3. Service Availability

Formula for calculating Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) /Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of Cloud CSS per Service Cycle not less than 99%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the following service credit:

Service   Availability Rate	Service   Credit
97% ≤ Service   Availability Rate < 99%	10% of the Monthly   Service Fee
95%≤ Service   Availability Rate < 97%	20% of the Monthly   Service Fee
Service   Availability Rate < 95%	50% of the Monthly   Service Fee

Last Updated: January 15, 2024

Cloud Data Migration (CDM) Service Specific Terms

1. Definitions

Service Unavailability means that according to the CDM system log, CDM was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Customer operations, such as deleting, stopping, or restarting clusters, or deleting jobs, connections, or drivers through the console or APIs;

2) Planned system maintenance, including but not limited to cutover, maintenance, upgrade, patching, and fault simulations, of which HUAWEI CLOUD notifies the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Hacker attacks on user applications;

5) Consequences of data leakage (including passwords) or loss due to the fault of the customer;

6) Operations authorized by the customer;

7) Customer's failure to comply with the product usage documentation or recommendations provided by HUAWEI CLOUD;

8) Scheduling failures or failures to start scheduling caused by improper job scheduling configurations;

9) Job execution failures caused by password or credential expiration of the account configured by the customer;

10) Job execution failures caused by insufficient resources such as the CPU, memory, disk I/O, and network I/O of the CDM cluster when customer services are overloaded;

11) Job execution failures caused by the unavailability of source or destination data sources.

3. Service Availability

Formula to calculate Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of CDM per Service Cycle is not less than 99.9%.

4. Service Credit

If we fail to meet the above commitment, we will provide you with the following service credit.

Service Availability Rate	Service Credit
99.0% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate < 99.0%	20% of Monthly Service Fee

Last Updated: January 15, 2024

NAT Gateway（NAT）Service Specific Terms

1. Definitions

Service Cycle means a calendar month.

Total Time of Service Cycle means the total minutes during every Service Cycle.

Service Unavailability means all outgoing data packets of a NAT gateway are discarded within one minute. The Service Unavailable Duration is the total number of minutes in a given service cycle that all outgoing packets of a NAT gateway were discarded.

Monthly Service Fee means service fees paid by customers for a NAT gateway in a Service Cycle (a calendar month). For example, if the customer pays the service fee of several months at a time, the Monthly Service Fee is calculated as the total service fees paid divided by the total number of months.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Attacks on your applications, including but not limited to DDoS attacks;

2) Faulty backend servers;

3) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

4) Network faults, device faults, or configuration adjustment of non-Huawei devices;

5) Loss or disclosure of data, including passwords, due to your improper maintenance;

6) Your negligence or allowed operations;

7) Your failure to follow the NAT Gateway documentation or usage suggestions, or by your refusal to take suggestions after being notified of risks;

8) Your violation against applicable laws and regulations, requirements of relevant authorities, or rules of Huawei Cloud platform;

9) Flow control over requests due to your service overload;

10) Force majeure or other accidents that are unforeseeable, insurmountable, and unavoidable and have significant impact on one party or both parties, including but not limited to natural disasters (such as floods, earthquakes, and plague epidemics) and social events (such as wars, unrest, government actions, interruption of telecom backbone lines, hackers, network congestion, technical change of telecommunications departments, and government policies).

3. Service Availability

Formula for calculating Service Availability in each Service Cycle

Service Availability in each Service Cycle = (Total Time of Service Cycle - Service Unavailable Duration) / Total Time of Service Cycle x 100%.

Service Availability Commitment

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability of NAT Gateway per Service Cycle is not less than 99.95%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the below service credit:

Service Availability	Service Credit
99.00% ≤ Service Availability < 99.95%	10% of Monthly Service Fee
95.00% ≤ Service Availability < 99.00%	25% of Monthly Service Fee
Service Availability < 95.00%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Virtual Private Network (VPN) Service Specific Terms

1. Definitions

Service Unavailability means all outgoing data packets of a VPN gateway are discarded for at least one minute. The Service Unavailable Duration is the number of minutes in a given service cycle during which all packets were discarded. Service unavailability durations shorter than one minute are not included.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Data cannot be sent due to an IPsec tunnel negotiation failure;

2) A Huawei Cloud VPN gateway has sent data out over the Internet, but the data fails to be received due to a failure of your public network;

3) For Enterprise Edition VPN, only one VPN connection is created between an Enterprise Edition VPN gateway and a customer gateway.

3. Service Availability

Calculation formula for the Service Availability Rate in each Service Cycle:

Service Availability Rate (%) in a Service Cycle = (Total Time of Service Duration – Service Unavailable Duration in this Service Cycle)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of VPN per Service Cycle is not less than 99.95%.

4. Service Credit

If we fail to meet these commitments, HUAWEI CLOUD will provide you with the service credit described below:

Service   Availability Rate	Service Credit
99%   ≤ Service Availability Rate < 99.95%	10%   of Monthly Service Fee
95%   ≤ Service Availability Rate < 99%	25%   of Monthly Service Fee
Service   Availability Rate < 95%	100%   of Monthly Service Fee

Last Updated: January 15, 2024

Media Processing Center (MPC) Service Specific Terms

1. Definitions

Failed Requests are requests with 5XX status codes, normal requests that fail to reach the MPC server due to a service fault, or requests that have not been processed in 24 hours after being submitted.

Valid Requests are requests received by the MPC server or normal requests that fail to reach the MPC server due to a service fault. Requests with 403 status code, such as ones that fail identity identification or authentication, or ones initiated using accounts with overdue amounts are not valid requests.

Error Rate in an Interval

Error rate in an interval = Number of failed requests in the interval/Number of valid requests in the interval x 100%

Statistics are collected every 5 minutes, and 288 statistics points are collected every day.

Service Unavailability: If Error rate in an interval > 0.05%, MPC is unavailable.

2. Disclaimers

Service Unavailability caused by the following situations shall not be counted in the Service Unavailable Duration:

1) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which HUAWEI CLOUD has notified the customer in advance;

2) Sharp increase in transcoding volume, which causes the daily transcoding output duration to increase by more than 100,000 minutes.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (1 – Service Unavailable Duration in Service Cycle/Total Time of Service Cycle) x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of MPC per Service Cycle no less than 99.9%.

4. Service Credit

In the event that MPC does not meet the above Commitment, HUAWEI CLOUD will provide you with the below service credit. However, the total compensation amount shall not exceed 25% of the Monthly Service Fee (excluding the fee deducted using cash coupons) you paid.

Service Availability	Service Credit
99% ≤ Service Availability Rate <   99.9%	10% of Monthly Service Fee
Service Availability Rate < 99%	25% of Monthly Service Fee

Last Updated: January 15, 2024

API Gateway (APIG) Service Specific Terms

1. Definitions

Service Unavailable means the API Gateway log shows that API Gateway was inaccessible or all API requests failed for more than one minute due to the fault of HUAWEI CLOUD.

Valid Requests means all requests received by API Gateway.

Failed Requests means API requests that failed due to the fault of API Gateway, excluding requests that failed due to the following causes:

1) The API configuration or dedicated gateway configuration (such as unbinding and deleting an EIP) is incorrect;

2) The API backend service is abnormal due to response timeout, connection failure, network failure, or service logic error;

3) The number of requests using the API group subdomain name automatically generated by API Gateway has reached the maximum daily limit;

4) The user's application is maliciously attacked;

5) The number of API requests has reached the maximum limit allowed by API Gateway specifications;

6) The user account is in arrears or is suspended;

7) Incorrect configurations made by the user.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate no less than 99.9% for the shared gateway, no less than 99.95% for the dedicated gateway basic, professional, and enterprise editions (multi-AZ deployment), and no less than 99.99% for the dedicated gateway platinum edition (multi-AZ deployment).

3. Service Credit

If we fail to meet the commitment above on API Gateway Service Availability Rate, HUAWEI CLOUD will provide you with the following service credits:

Shared gateway

Service Availability   Rate	Service Credit
99.0% ≤ Service Availability Rate < 99.9%	15% of Monthly Service Fee
95.0% ≤ Service Availability Rate < 99.0%	30% of Monthly Service Fee
Service Availability Rate < 95.0%	100% of Monthly Service Fee

Dedicated gateway (basic, professional, enterprise, and platinum editions; single-AZ deployment)

Service Availability   Rate	Service Credit
99.0% ≤ Service Availability Rate < 99.9%	15% of Monthly Service Fee
95.0% ≤ Service Availability Rate < 99.0%	30% of Monthly Service Fee
Service Availability Rate < 95.0%	100% of Monthly Service Fee

Dedicated gateway (basic, professional, and enterprise editions; multi-AZ deployment)

Service Availability   Rate	Service Credit
99.0% ≤ Service Availability Rate < 99.95%	15% of Monthly Service Fee
95.0% ≤ Service Availability Rate < 99.0%	30% of Monthly Service Fee
Service Availability Rate < 95.0%	100% of Monthly Service Fee

Dedicated gateway (platinum edition; multi-AZ deployment)

Service Availability   Rate	Service Credit
99.0% ≤ Service Availability Rate < 99.99%	15% of Monthly Service Fee
95.0% ≤ Service Availability Rate < 99.0%	30% of Monthly Service Fee
Service Availability Rate < 95.0%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Data Security Center (DSC) Service Specific Terms

1. Definitions

Service Unavailability: The DSC system log shows that DSC was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD. Duration of the inaccessibility and incapability that lasts shorter than five (5) minutes shall not be included.

2. Service Availability

Calculation formula of service availability rate in each service cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD promises a Service Availability Rate of up to 99.9% for DSC through substantial commercial and technical effort.

3. Service Credit

In the event HUAWEI CLOUD does not meet the commitment on the Service Availability Rate specified in this SLA, you will be eligible to receive service credit as described below:

Service Availability Rate	Service   Credit
99% ≤   Service Availability Rate < 99.9%	15% of monthly service fee of the involved faulty   service or instance
95% ≤   Service Availability Rate < 99%	30% of monthly service fee of the involved faulty   service or instance
Service   Availability Rate < 95%	100% of monthly service fee of the involved   faulty service or instance

Last Updated: January 15, 2024

Database Security Service (DBSS) Service Specific Terms

1. Definition

Service Unavailability means the DBSS system log shows that DBSS instances were inaccessible for 5 or more consecutive minutes due to HUAWEI CLOUD.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the DBSS Service Availability Rate in each Service Cycle no less than 99.95%.

3. Service Credit

If we fail to meet the commitment above on DBSS Service Availability Rate, HUAWEI CLOUD will provide you with the following service credits: 

Service   Availability Rate	Service Credit
99.00%   ≤ Service Availability Rate < 99.95%	15%   of Monthly Service Fee
95.00%   ≤ Service Availability Rate < 99.0%	30%   of Monthly Service Fee
Service   Availability Rate < 95.00%	100%   of Monthly Service Fee

Last Updated: January 15, 2024

Key Management Service (KMS) Service Specific Terms

1. Definition

Service Unavailability means the KMS system log shows that KMS clusters were inaccessible for 5 or more consecutive minutes due to HUAWEI CLOUD.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the KMS Service Availability Rate in each Service Cycle no less than 99.95%.

3. Service Credit

If we fail to meet the above commitment, HUAWEI CLOUD will provide you with the below service credit:

Service   Availability Rate	Service Credit
99.00%   ≤ Service Availability Rate < 99.95%	15%   of Monthly Service Fee
95.00%   ≤ Service Availability Rate < 99.0%	30%   of Monthly Service Fee
Service   Availability Rate < 95.00%	100%   of Monthly Service Fee

Last Updated: January 15, 2024

Web Application Firewall (WAF) Service Specific Terms

1. Definitions

Service Unavailability: System logs show that the WAF service was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD.

2. Disclaimers

For WAF, Service Unavailability in the following conditions cannot be counted in to Service Unavailable Duration, in addition to the conditions described in the definition of Service Unavailability:

1) Your backend origin servers are abnormal due to certain reasons, for example, bandwidth exhaustion, origin IP address exposure, data center breakdowns, network jitters, installation or upgrade of OSs or applications;

2) Your website requests exceeds the concurrent requests that can be handled by the WAF edition you are using;

3) There are attacks not at the web application layer, for example, heavy DDoS attacks;

4) Data such as passwords is lost or leaked due to your improper maintenance.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Minutes in the Service Cycle – Unavailable Minutes in the Service Cycle)/Total Minutes in the Service Cycle) x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make commercially viable and technically feasible efforts to ensure that the WAF Service Availability Rate per Service Cycle is not less than 99.95%.

4. Service Credit

If WAF fails to meet the commitment on the service availability rate, HUAWEI CLOUD will compensate you based on the terms given in the following table.

Service   Availability Rate	Coupon
99.00% ≤ Service Availability   Rate < 99.95%	10% of the basic monthly payment
95.00% ≤ Service Availability   Rate < 99.95%	25% of the basic monthly payment
Service Availability Rate < 95%	100% of the basic monthly fee

Last Updated: January 15, 2024

HUAWEI CLOUD Meeting (Meeting) Service Specific Terms

1. Definitions

Service Unavailability means that According to logs in the Huawei Cloud Meeting, meeting cannot be held in any way for more than 5 consecutive minutes due to Huawei Cloud Meeting reasons. Unavailability time less than 5 minutes is not counted.

2. Service Availability

Calculation formula of service availability rate in each service cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud Meeting will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of Huawei Cloud Meeting per Service Cycle is not less than 99.9%.

3. Service Credit

If we fail to meet these commitments, Huawei Cloud Meeting will provide you with the service credit described below: 

Service Availability	Service Credit
99 % ≤ Service   Availability Rate < 99.9%	5% of monthly service fee
97% ≤ Service   Availability Rate < 99 %	10% of monthly service fee
95% ≤ Service   Availability Rate < 97%	20% of monthly service fee
Service Availability Rate   < 95%	50% of monthly service fee

Last Updated: January 15, 2024

Container Guard Service (CGS) Service Specific Terms

1. Definition

Service Unavailability means the CGS system log shows that CGS instances were inaccessible for 5 or more consecutive minutes due to HUAWEI CLOUD.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in Each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the CGS Service Availability Rate in each Service Cycle no less than 99.9%.

3. Service Credit

If we fail to meet the commitment above on CGS Service Availability Rate, HUAWEI CLOUD will provide you with the following service credits:

Service Availability Rate	Service Credit
99% ≤ Service   Availability Rate < 99.9%	10% of Monthly   Service Fee
Service   Availability Rate < 99%	25% of Monthly   Service Fee

Last Updated: January 15, 2024

Situation Awareness (SA) Service Specific Terms

1. Definitions

Service Unavailability means the SA system log shows that SA was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

A monthly service uptime percentage calculated based on the following formula. SA of each individual subscriber is separately measured.

Monthly Service Availability Rate = [Service availability duration in minutes of a billing month/Total number of minutes of a billing month] x 100%.

If the duration is less than one month, it is still regarded as one month.

Service Availability Rate commitment:

HUAWEI CLOUD shall make commercially and technically reasonable efforts to make the Service Availability Rate for SA up to 99.9%. Should HUAWEI CLOUD fail to ensure the promised rate, HUAWEI CLOUD will compensate you in accordance with Clause 3 of this Agreement.

3. Service Credit

If SA fails to meet the commitment on the service availability rate, HUAWEI CLOUD will compensate you based on the terms given in the following table.

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate < 99%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Distributed Database Middleware (DDM) Service Specific Terms

1. Definition

Service Unavailability means that the DDM system log shows that DDM instances were inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD. Duration of the inaccessibility and inability that lasts for shorter than one (1) minute shall not be included.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by:

1) Customer operations, such as restarting, deleting, or restoring DDM instances through a console or APIs;

2) Network faults, device faults, or configuration adjustment of non-Huawei devices;

3) Hacker attacks on customer applications;

4) Planned system maintenance, including but not limited to cutover, maintenance, upgrade, and fault drills, of which HUAWEI CLOUD has notified customers in advance;

5) Changing or deleting accounts of databases in the RDS DB instances that are associated with the DDM instance;

6) Deleting data stored in databases in the RDS DB instances that are associated with the DDM instance;

7) Modifying or deleting the VPC virtual switch used by the DDM instance;

8) Data (including passwords) leakage or loss due to improper maintenance of the customer;

9) Operations authorized by the customer or any misjudgment in operations;

10) Customer's failure to comply with the DDM user guide or recommendations;

11) Service interruptions caused by insufficient vCPUs, memory, storage, or the maximum number of connections being reached for a DB instance or a database proxy due to heavy customer workloads.

3. Service Availability

Formula for calculating Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make commercially and technically reasonable efforts to ensure that the Service Availability Rate per Service Cycle is not less than 99.90%.

4. Service Credit

If we fail to meet the above commitment, HUAWEI CLOUD will provide you with service credit as follows.

Service    Availability	Service    Credit
99% ≤ Service Availability Rate <   99.90%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: March 20, 2024

Cloud Connect (CC) Service Specific Terms

1. Definition

Service Unavailable means that network instances cannot communicate with each other over the cloud connection that connects them for one 1 or more consecutive minutes. Service Unavailable Duration refers to the total minutes in a Service Cycle in which network instances cannot communicate with each other over the cloud connection.

2. Disclaimers

This SLA does not apply to service unavailability caused by Direct Connect or VPN even if a Direct Connect or VPN gateway is connected to a cloud connection to allow your on-premises data center to access other network instances connected through the cloud connection.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make commercially and technically reasonable efforts to ensure that the Cloud Connect service availability is not lower than 99.95%.

4. Service Credit

If we fail to meet the commitment above on the Cloud Connect Service Availability Rate, HUAWEI CLOUD will compensate you as showed in the following table:

Service   Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.95%	15% of Monthly Service Fee
95.00% ≤ Service Availability Rate < 99.00%	30% of Monthly Service Fee
Service Availability Rate < 95.00%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Simple Message Notification (SMN) Service Specific Terms

1. Definition

Service Unavailability means the SMN system log shows that SMN was inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration *100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make all commercially viable and technically feasible efforts to ensure that the Service Availability Rate of SMN per Service Cycle is not less than 99.9%.

3. Service Credit

If we fail to meet these commitments, HUAWEI CLOUD will provide you with the service credit described below:

Service   Availability Rate	Service   Credit
95.00% ≤Service Availability Rate   < 99.90%	10% of Monthly Service Fee
Service Availability Rate <   95.00%	20% of Monthly Service Fee

Service Unavailability does not include unavailability resulting from any of the following issues:

1) Planned system maintenance, such as cutovers, maintenance, upgrades, or fault simulations that HUAWEI CLOUD has notified the customer of in advance;

2) Network faults, device faults, or configuration adjustments of non-HUAWEI CLOUD devices;

3) Online attacks on customer applications;

4) Loss or disclosure of data, including passwords, due to improper maintenance by the customer;

5) Operations authorized by the customer or any mis-judgment in operations;

6) Major events or promotions that HUAWEI CLOUD has notified the customer of in advance;

7) Customer violations of the information control, traffic control, or information security control rules of HUAWEI CLOUD or other carrier platforms;

8) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

9) Customer violations of local laws or regulations;

10) Messages from SMN to the customer were blocked;

11) The number of messages sent by the customer exceeds the maximum allowed number of messages specified in the Simple Message Notification API Reference and the customer did not notify HUAWEI CLOUD in advance;

12) Any force majeure event.

Last Updated: January 15, 2024

GaussDB Service Specific Terms

1. Definition

Service Unavailability: The GaussDB system log shows that a GaussDB instance was inaccessible for one (1) or more consecutive minutes due to a fault on Huawei Cloud.

2. Disclaimers

Service unavailability caused by the following situations shall not be counted as part of the service unavailability duration:

1) Customer operations, such as rebooting, deleting, or restoring GaussDB instances through the console or APIs;

2) Network faults, device faults, or configuration adjustment of non-Huawei Cloud devices;

3) Hacker attacks on customer applications;

4) Planned system maintenance, such as migration, maintenance, upgrade, or fault drills, of which Huawei Cloud has notified the customer in advance;

5) Data (including passwords) leakage or loss due to improper maintenance of the customer;

6) Operations authorized by the customer or any misjudgment in operations;

7) Non-compliance with GaussDB user guide, development specifications, or usage suggestions;

8) Service interruption caused by insufficient vCPUs or storage due to heavy customer workloads;

9) Faults caused by open-source community bugs or other non-Huawei Cloud issues;

10) Service suspension or termination due to a breach of service on the customer's end (such as an unpaid account balance);

11) Customer's failure to rectify known database problems or risks which Huawei Cloud has notified;

12) Faults caused by single-replica instances;

13) Time consumed during log replay (including log redo or recovery) of instances;

14) Issues caused by improper service use that deviates from the instructions provided in documents such as Best Practices and FAQs;

15) Customer operation to choose the deployment that cannot ensure the high availability of GaussDB instances;

16) Issues introduced by third-party open-source software;

17) Other force majeure factors.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will make all reasonable commercial and technical efforts to ensure:

The service availability rate of the following instances per service cycle is at least 99.99%: three-AZ instances (1 primary + 2 standby).

The service availability rate of the following instances per service cycle is at least 99.95%: single- and two-AZ instances (1 primary + 2 standby) and instances (1 primary + 1 standby +1 log).

4. Service Credit

If we fail to meet this commitment, we will provide you with the following service credits:

Three-AZ instances (1 primary + 2 standby)

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.99%	10% of Monthly Service Fee
95.00% ≤ Service Availability Rate < 99.00%	25% of Monthly Service Fee
Service Availability Rate < 95.00%	100% of Monthly Service Fee

Single- and two-AZ instances (1 primary + 2 standby)/Instances (1 primary + 1 standby +1 log):

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
95.00% ≤ Service Availability Rate < 99.00%	25% of Monthly Service Fee
Service Availability Rate < 95.00%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Managed Threat Detection (MTD) Service Specific Terms

1. Definition

HUAWEI CLOUD MTD service continuously monitors malicious activities and unauthorized behaviors to protect accounts and workloads. The service scope depends on your purchase and the services provided by HUAWEI CLOUD.

2. Service unavailable duration

1) MTD is suspended for more than 5 consecutive minutes, with the downtime of less than 5 minutes not being counted;

2) Service suspension caused by account renewal is not counted;

3) Service suspension compliant with user agreements is not counted;

4) Service suspension caused by Internet network faults is not counted;

5) Service suspension caused by content or applications attacked or infected with viruses in the network;

6) Service suspension caused by failure to comply with user agreements and laws and regulations.

3. Service availability

Formula calculating the service availability rate in each service cycle

Service availability rate in each service cycle = (Total time of service duration – Service unavailable duration) / Total time of service duration x 100%.

Service availability rate

HUAWEI CLOUD will use commercial and technical methods to ensure that MTD availability is not lower than 99.9%.

4. Service credit

If the MTD fails to fulfill commitments on the service availability rate, the service credit amount shall be calculated according to the following table:

Service   Availability Rate	Service   Credit
99%   ≤ Service availability rate < 99.9%	10%   of monthly service fee
Service   availability rate < 99%	25%   of monthly service fee

Last Updated: January 15, 2024

Distributed Message Service (DMS) Service Specific Terms

1. Definitions

Service Unavailability means the DMS system log shows that DMS was inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD. Service Unavailability does not include unavailability resulting from any of the following issues:

1) Network faults, device faults, or configuration adjustment of non-HUAWEI CLOUD devices;

2) Issues caused by improper product usage, such as overloaded broker, insufficient disk space, and lack of multiple replicas or redundant queues;

3) Faults that occur due to single-AZ deployment, instead of multi-AZ deployment;

4) Issues caused by third-party open-source software.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of DMS per Service Cycle not less than 99.9%.

3. Service Credit

If we fail to meet the above commitment, HUAWEI CLOUD will provide you service credit as shown in the following table.

Service   Availability	Service Credit
99% ≤ Service   Availability Rate < 99.9%	10% of Monthly   Service Fee
95% ≤ Service   Availability Rate < 99%	30% of Monthly   Service Fee
Service   Availability Rate < 95%	100% of Monthly   Service Fee

Last Updated: January 15, 2024

Distributed Cache Service (DCS) Service Specific Terms

1. Definition

Service Unavailability means the DCS system log shows that DCS (for Redis and for Memcached) was inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD. Service Unavailability does not include unavailability resulting from any of the following issues:

1) Network faults, device faults, or configuration adjustment of non-HUAWEI CLOUD devices;

2) Issues caused by improper service use (such as exceeding the CPU, memory, and network bandwidth limits of instance nodes) that deviates from the instructions provided in documents such as Best Practices and FAQs;

3) Faults that occur due to the use of single-node instances, instead of HA instances such as master/standby or cluster instances;

4) Faults that occur due to single-AZ deployment, instead of multi-AZ deployment;

5) Issues caused by third-party open-source software.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of DCS per Service Cycle not less than 99.95%.

3. Service Credit

If the above commitment fails to be met due to HUAWEI CLOUD reasons, HUAWEI CLOUD will provide you service credit as shown in the following table.

Master/standby Redis instances:

Service   Availability	Service Credit
99% ≤ Service Availability   Rate < 99.95%	10% of Monthly   Service Fee
95% ≤ Service   Availability Rate < 99%	30% of Monthly   Service Fee
Service   Availability Rate < 95%	100% of monthly   service fee

Cluster Redis instances:

Service   Availability	Service Credit
99% ≤ Service   Availability Rate < 99.95%	10% of Monthly   Service Fee
95% ≤ Service   Availability Rate < 99%	30% of Monthly   Service Fee
Service   Availability Rate < 95%	100% of monthly   service fee

Read/write splitting Redis instances:

Service   Availability	Service Credit
99% ≤ Service   Availability Rate < 99.95%	10% of Monthly   Service Fee
95% ≤ Service   Availability Rate < 99%	30% of Monthly   Service Fee
Service   Availability Rate < 95%	100% of monthly   service fee

Master/standby Memcached instances:

Service Availability	Service Credit
99% ≤ Service   Availability Rate < 99.95%	10% of Monthly   Service Fee
95% ≤ Service   Availability Rate < 99%	30% of Monthly   Service Fee
Service   Availability Rate < 95%	100% of monthly   service fee

Last Updated: January 15, 2024

Application Operations Management (AOM) Service Specific Terms

1. Definition

Service Unavailable means the AOM system log shows that AOM was inaccessible for more than five consecutive minutes due to a fault with HUAWEI CLOUD.

2. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure that the service availability rate is at least 99.9% for AOM in each service cycle.

3. Service Credit

If we fail to meet the commitments described here for the AOM service availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service   Availability Rate	Service   Credit
99% ≤ Service Availability Rate <   99.9%	10% of Monthly Service Fee
95% ≤ Service Availability Rate <   99%	30% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Blockchain Service (BCS) Service Specific Terms

1. Definition

Service Unavailable means the BCS system log shows that BCS was inaccessible for more than ten consecutive minutes due to a fault with HUAWEI CLOUD.

2. Disclaimers

Service Unavailable does not include unavailability resulting from the following issues:

1)  Logical defects of user smart contracts;

2) Misoperation and constraints of compute, storage, and network resources;

3) OBT and beta features.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure that the Service Availability Rate of BCS per Service Cycle is at least 99.95%.

4. Service Credit

If we fail to meet the commitment described here for the BCS Service Availability Rate, HUAWEI CLOUD will provide you with the service credit below.

Service   Availability Rate	Service   Credit
99% ≤ Service Availability   Rate < 99.95%	10% of Monthly Service Fee
Service Availability Rate   < 99%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Bare Metal Server (BMS) Service Specific Terms

1. Definition

Service Unavailable means the BMS system log shows that BMS was inaccessible for more than five consecutive minutes due to a fault with HUAWEI CLOUD.

2.Disclaimers

This SLA does not apply to any performance or availability issues:

1) caused by your negligence or HUAWEI CLOUD operations authorized by you on servers;

2) caused by network faults, device faults, or configuration adjustment of non-Huawei devices;

3) caused by server hardware faults you refused to authorize HUAWEI CLOUD to repair;

4) caused by hardware faults of servers whose warranty has expired and then you refused to upgrade them;

5) that persist when Huawei is waiting for your authorization for server hardware repair and when Huawei engineers are repairing it;

6) that occur on a single piece of hardware and do not affect your services due to hardware redundancy. For example, if a single NIC in a bond is faulty or a single disk in a storage redundancy group is faulty, and the faults do not affect your services, the SLA does not apply;

7) caused by local disks. Data loss may occur on local disks in the cases, such as if the disks are damaged or the BMSs break down;

8) caused by OS or key process issues when you install, reinstall, or uninstall your own or third-party applications;

9) caused by hacker attacks on your applications;

10) caused by data (including passwords) leaked or lost due to user errors;

11) caused by server disconnection resulted from normal server operations including but not limited to OS installation, OS reinstallation, BIOS modifications, server shutdowns, or server restarts;

12) that occur when you do not follow the HUAWEI CLOUD product documentation or other guidelines to perform operations such as stopping a BMS, restarting a BMS, and detaching EVS disks from a BMS using APIs, CLI, or console.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure the service availability rate is at least 99.9% for BMS in each service cycle.

4. Service Credit

If we fail to meet the commitments described here for the BMS service availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate < 99%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Cloud Container Engine (CCE) Service Specific Terms

1. Definition

Service Unavailability means the CCE system log shows that the Kubernetes native APIs in a CCE cluster were inaccessible in the way supported by CCE for 5 consecutive minutes or more due to Huawei Cloud.

2. Disclaimers

For CCE, Service Unavailability in the following conditions cannot be counted in to Service Unavailable Duration, in addition to the conditions described in the definition of Service Unavailability:

1) CCE service availability is undermined due to the risky and impactful operations CCE has explicitly notified you of, the execution of which may disable the cluster control plane to allow access or provide services within the declared scope and time;

2) Any undermined performance or availability due to exceptions such as the unavailability or resource shortage of its dependent services, including Huawei Cloud compute, storage, and network services, to which CCE connects to run Kubernetes clusters. In this case, the SLA of the related cloud services applies;

3) CCE service availability is undermined due to issues from any third party, including Kubernetes, third-party Kubernetes add-ons integrated with CCE, Docker, and OS kernel;

4) CCE service availability is undermined due to the failure of Kubernetes nodes or the Kubernetes pods running on those nodes;

5) The cluster fails due to your change on the configurations without following the CCE documentation or console instructions, unless otherwise specified;

6) CCE service SLA only applies to the CCE cluster commercial version, and does not apply to the CCE cluster OBT and EOS versions. For details, see Kubernetes Version Policy in the CCE bulletin;

7) You are using a feature that is not in the OBT stage as specified in the CCE documentation or console. In this case, CCE does not promise the SLA, unless otherwise specified;

8) CCE service availability is undermined due to acts, events, omissions, or incidents arising from events that are unforeseeable, inevitable, and beyond its reasonable control, including but not limited to natural disasters (earthquakes, storms, etc.), act of terrorism, war or warlike operations, civil unrest or riots, electricity outages, network or telecommunication faults, blockages, embargoes, fires, floods, explosions or malicious damage (including but not limited to hacker attacks), plant or equipment failure, or changes of any law, government order, rule, regulation, direction, or industry standard.

3. Service Availability (SA)

SA in each Service Cycle is represented by the following formula:

SA in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

SA commitment:

Huawei Cloud makes reasonable commercial and technical efforts to ensure the SA of a cluster in each service cycle as defined below:

• SA of an HA cluster (3-AZ) ≥ 99.95%
• SA of an HA cluster (1-AZ or 2-AZ) ≥ 99.9%
• SA of a non-HA cluster ≥ 99.5%

4. Service Credit

If we fail to meet the commitments described here for the CCE service availability rate, Huawei Cloud will provide you with the service credit below.

HA cluster (3-AZ)

Service Availability (SA)	Coupon Amount
99% ≤ SA < 99.95%	10% of monthly service fee
95% ≤ SA < 99%	25% of monthly service fee
SA < 95%	100% of monthly service fee

 HA cluster (1-AZ or 2-AZ)

Service Availability (SA)	Coupon Amount
99% ≤ SA < 99.9%	10% of monthly service fee
95% ≤ SA < 99%	25% of monthly service fee
SA < 95%	100% of monthly service fee

Non-HA cluster

Service Availability (SA)	Coupon Amount
99% ≤ SA < 99.5%	10% of monthly service fee
95% ≤ SA < 99%	25% of monthly service fee
SA < 95%	100% of monthly service fee

 Last Updated: May 8, 2024

 

You can see what is updated in Cloud Container Engine (CCE) Service Level Agreement History Version.

Content Moderation (Moderation) Service Specific Terms

1. Definition

Service Unavailable means the Content Moderation system log shows that Content Moderation was inaccessible for more than five consecutive minutes due to a fault with HUAWEI CLOUD.

2. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure the service availability rate is at least 99.9% for Content Moderation in each service cycle.

3. Service Credit

If we fail to meet the commitments described here for the Content Moderation service availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service   Availability Rate	Service   Credit
99.0% ≤ Service   Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate   < 99.0%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Cloud Search Service (CSS) Service Specific Terms

1. Definition

Service Unavailability means the CSS system log shows that the CSS cluster was inaccessible for five consecutive minutes or more due to Huawei Cloud.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

A monthly service uptime percentage calculated based on the following formula. CSS of each individual subscriber is separately measured. Monthly Service Availability Rate = (Service Availability Duration in Minutes of a Billing Month/Total Number of Minutes of a Billing Month) x 100%. A billing month that is less than one (1) month shall be counted as a full month.

Service Availability Rate Commitment:

Huawei Cloud will use all commercially and technically reasonable efforts to ensure the service availability rate of CSS is at least 99.95% for the service plane (for cluster access), and at least 99.9% for the management plane (for cluster addition, deletion, modification, and query). If we fail to meet the commitments described here for the CSS service availability rate, Huawei Cloud will compensate you with coupons for the amounts shown in the following table.

3. Service Credit

If we fail to meet the commitment above on CSS Service Availability Rate, Huawei Cloud will provide you service credit as shown in the following table:

Service Availability Rate	Service Credit
99.9% ≤ Service Availability Rate < 99.95%	5% of Monthly Service Fee
99% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate < 99%	25% of Monthly Service Fee

4. Restriction

This SLA does not apply to any performance or availability issues caused by the following:

1) The cluster has fewer than 3 nodes;

2) You incorrectly configure cluster parameters;

3) Your service has been overloaded, but you continue to make frequent data read/write or changes;

4) You do not enable replicas as suggested for your cluster that uses local disks. As a result, data cannot be restored in the case of a data loss;

5) You do not enable 3-AZ HA for your cluster, and your cluster is unavailable due to an AZ fault;

6) Known limitations or issues of the open-source underlying software.

The CSS SLA only applies to clusters of the CSS commercial version. It does not apply to clusters of the CSS OBT version or EOS version. 

Last Updated: June 7, 2024

You can see what is updated in Cloud Search Service (CSS) Service Level Agreement History Version.

DataArts Studio Service Specific Terms

1. Definitions

Service Unavailability means that according to the DataArts Studio system log, DataArts Studio was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Customer operations, such as deleting, stopping, or restarting clusters, or deleting jobs, connections, or drivers through the console or APIs;

2) Planned system maintenance, including but not limited to cutover, maintenance, upgrade, patching, and fault simulations, of which HUAWEI CLOUD notifies the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Hacker attacks on user applications;

5) Consequences of data leakage (including passwords) or loss due to the fault of the customer;

6) Operations authorized by the customer;

7) Customer's failure to comply with the product usage documentation or recommendations provided by HUAWEI CLOUD;

8) Scheduling failures, failures to start scheduling, job execution timeout, or failures to retry jobs upon execution failures caused by improper job scheduling configurations;

9) Job execution failures caused by password or credential expiration of the account configured by the customer.

3. Service Availability

Formula to calculate Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of DataArts Studio per Service Cycle is not less than 99.95%.

4. Service Credit

If we fail to meet the above commitment, we will provide you with the following service credit.

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
Service Availability Rate < 99.00%	25% of Monthly Service Fee

Last Updated: January 15, 2024

IoT Edge Service Level Agreement

1. Definition

Service Unavailable means the IoT Edge system log shows that IoT Edge was inaccessible for more than five consecutive minutes due to a fault with HUAWEI CLOUD.

2. Disclaimers

This SLA does not apply to any performance or availability issues:

1) The selected edge hardware (such as the OS, hardware architecture, memory, hard disk, and CPU) does not meet the usage constraints specified by IoT Edge;

2) Edge nodes and edge applications cannot be managed by IoT Edge due to network faults (for example, edge node network disconnection), hardware faults (for example, hard disk damage), OS faults, and Docker software faults;

3)  When an edge node is offline, the O&M information (including logs, alarms, and monitoring information) of edge nodes and edge applications cannot be reported to the cloud. In this case, IoT Edge cannot manage the edge nodes and edge applications;

4) Edge nodes and edge applications are abnormal due to system faults caused by the deployed edge applications (for example, the edge applications are overloaded or the disk space is used up);

5) Services are interrupted or unavailable because you do not properly plan the edge application upgrade policy;

6) Services are interrupted or unavailable due to the OS upgrade, Docker software upgrade, or IoT edge software upgrade of your edge node.

Version Maintenance Interval: The IoT edge software you use must be upgraded to the latest version within six months to ensure performance and receive the O&M support provided by the IoT Edge service team. After the maintenance interval (six months), the service team does not provide free O&M support for your IoT edge software.

O&M Responsibility Division:

1) HUAWEI CLOUD is responsible for O&M of IoT edge software;

2) You are responsible for routine O&M of edge nodes and edge applications.

a)Hardware devices, OSs, and networks of edge nodes are maintained by you or your project system integrator.

b)Edge applications are maintained by you or your project system integrator.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure the service availability rate is at least 99% for IoT Edge in each service cycle.

4. Service Credit

If we fail to meet the commitments described here for the IoT Edge service availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service   Availability Rate	Service   Credit
95% ≤ Service Availability   Rate < 99%	10% of Monthly Service Fee
Service Availability Rate   < 95%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Live Service Specific Terms

1. Definition

Service Cycle means a natural month.

Failed Requests means API requests that failed due to a fault with Live. They include FLV-HTTP or HLS requests with 5XX status codes and RTMP requests whose response is not returned within 20 seconds.

Requests that failed due to the following causes are excluded:

1) Playback failed because signed URLs expire when URL validation is enabled;

2) Domain names are banned due to content violation or other reason;

3) Playback failed due to exceptions on streaming devices;

4) Requests are incorrect or services are unavailable caused by reasonable upgrade, change, or suspension initiated by Live.

Total Time of Service Duration means the total number of minutes in each Service Cycle. Total Time of Service Duration = Total number of days in each Service Cycle x 24 hours x 60 minutes.

Service Unavailable Duration means the total duration that Live is unavailable during a Service Cycle. Service Unavailable Duration will be reset at the beginning of each Service Cycle.

Service Availability means the percentage of the time when Live is available in a Service Cycle.

Error rate in an interval of 5 minutes = Number of failed requests in 5 minutes / Total number of requests in 5 minutes x 100%.

Service Unavailable means that in an interval of 5 minutes, the number of requests is greater than 100 and the error rate calculated based on the data recorded in Live logs is greater than 0.05% due to a fault with HUAWEI CLOUD.

2. Disclaimers

If your Live bandwidth is expected to increase sharply (by at least 30% of the bandwidth billed last month), you shall notify Live by service ticket at least three working days in advance. Otherwise, any service unavailability will not be compensated.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure the Live Service Availability Rate is at least 99.9% per Service Cycle.

4. Service Credit

If we fail to meet the commitment described here for the Live Service Availability Rate, HUAWEI CLOUD will provide you with the service credit below (valid only for purchase of Live products). The total service credit shall not exceed 25% of Monthly Service Fee (excluding the fee deducted using coupons).

Service   Availability Rate	Service   Credit
99% ≤ Service Availability   Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate   < 99%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Log Tank Service (LTS) Service Specific Terms

1. Definition

Service Unavailable means the LTS system log shows that LTS was inaccessible for more than five consecutive minutes due to a fault with HUAWEI CLOUD.

2. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use all commercially and technically reasonable efforts to ensure a Service Availability Rate of LTS per Service Cycle of at least 99.9%.

3. Service Credit

If we fail to meet the commitments described here for the LTS service availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service   Availability Rate	Service   Credit
99% ≤ Service Availability   Rate < 99.9%	10% of Monthly Service Fee
95% ≤ Service Availability   Rate < 99%	30% of Monthly Service Fee
Service Availability Rate   < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

MapReduce Service (MRS) Service Specific Terms

1. Definitions

Service Unavailable means the MRS system log shows that MRS was inaccessible for more than five consecutive minutes due to a fault with Huawei Cloud.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Customer operations, such as deleting, stopping, or restarting clusters, or deleting jobs, connections, or drivers through the console or APIs;

2) Planned system maintenance, including but not limited to cutover, maintenance, upgrade, patching, and fault simulations, of which Huawei Cloud has notified the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Hacker attacks on user applications;

5) Consequences of data (including passwords) leakage or loss due to the fault of the customer;

6) Operations authorized by the customer;

7) Customer's failure to comply with the product usage documentation or recommendations provided by Huawei Cloud;

8) Customer's requests restricted by flow control due to quota exhaustion;

9) Service interruption caused by an unavailable OBS pool selected by the customer;

10) Customer's failure to deploy the cluster in HA mode (one Master node and one Core node, single  az deployment);

11) Data loss caused by the customer's failure to store multiple data replicas;

12) Customer's unauthorized deployment of applications and data on cluster nodes;

13) Other customer factors, for example, failure to execute customer-developed jobs, failure caused by the deployment of private components or the modification of system configurations/ privileges;

14) Other force majeure factors.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud strives to take commercially viable and technically feasible measures to ensure the Service Availability Rate of MRS per Service Cycle is at least 99.9%.

4. Service Credit

If we fail to meet the commitments described here for the MRS availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service Availability Rate	Service Credit
99.0% ≤ Service Availability Rate <   99.9%	10% of   Monthly Service Fee
Availability Rate < 99.0%	30% of Monthly   Service Fee

Last Updated: January 15, 2024

Optical Character Recognition (OCR) Service Specific Terms

1. Definition

Service Unavailability means the OCR system log shows that OCR was inaccessible for five (5) or more consecutive minutes because of HUAWEI CLOUD.

2. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will make all commercially and technically reasonable efforts to ensure the service availability rate is at least 99.9% for OCR in each service cycle.

3. Compensation

If we fail to meet the commitments described here for the OCR service availability rate, HUAWEI CLOUD will provide you with the service credit below.

Service   Availability Rate	Service   Credit
99.00% ≤ Service Availability   Rate < 99.90%	10% of Monthly Service Fee
Service Availability Rate <   99.00%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Whole Site Acceleration (WSA) Service Specific Terms

1. Definitions

Number of 5xx response errors in five minutes: Number of 5xx errors returned by the domain name due to WSA system problems in five minutes.

Number of failed requests due to node unavailability in five minutes: If a node is unavailable, calculate the number of requests per five minutes based on the statistics of the first seven days, and then convert it to the number of failed requests due to node unavailability in five minutes.

Error in five minutes = (Number of 5xx response error in five minutes + Number of failed requests due to node unavailability in five minutes) / Total number of requests in five minutes of the domain name

"Service Unavailability" means the WSA system log shows that the error rate of WSA was greater than 0.05% within five (5) minutes because of Huawei Cloud.

2. Disclaimers

If you will have a sharp increase in WSA bandwidth usage (greater than or equal to 30% of the billed bandwidth for last month), you must submit a service ticket to inform us at least three (3) business days in advance. Otherwise, HUAWEI CLOUD will not be responsible for service unavailability resulting from this.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of WSA per Service Cycle not less than 99.9%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the below service credit:

Service Availability	Service Credit
99%≤Service Availability Rate <99.9%	10%
Service Availability Rate <99%	25%

Last Updated: January 15, 2024

Scalable File Service  (SFS) Service Specific Terms

1. Definitions

Service Unavailability means the SFS system log shows that SFS was inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of SFS per Service Cycle not less than 99.95%.

3. Service Compensation

If we fail to meet the above commitment, we will provide you the service credit in the following table:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	30% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Enterprise Switch (ESW) Service Specific Terms

1. Definitions

Service Unavailability is any time data packets of an enterprise switch are discarded for one (1) or more consecutive minutes. The Service Unavailability Duration is the number of minutes in a given service cycle that all packets were discarded.

2. Disclaimers

Service Unavailability does not include unavailability resulting from any of the following:

1) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

2) Non-Huawei Cloud network failures, device faults, or configuration changes, including but not limited to service unavailability caused by network carrier line faults;

3) Application attacks including but not limited to DDoS attacks, SYN flood attacks, and virus attacks, or software and hardware compatibility issues;

4) Faulty backend servers;

5) Loss or disclosure of data, including passwords, due to your improper maintenance;

6) Operations authorized by you or any mis-judgment in operations;

7) Your failure to follow the product usage documentation or recommendations provided by Huawei Cloud;

8) Force majeure or other accidents that are unforeseeable, insurmountable, and unavoidable and have significant impact on one party or both parties, including but not limited to natural disasters (such as floods, earthquakes, and plague epidemics) and social events (such as wars, unrest, government actions, interruption of telecom backbone lines, hackers, network congestion, technical change of telecommunications departments, and government policies);

9) Network faults of other services (such as the customer's on-premises network fault, on-premises device fault, unavailable VPN, Direct Connect, and VPC);

10) Your services exceeding the supported resource specifications.

3. Service Availability

Service Availability in each Service Cycle is represented by the following formula:

Service Availability in each Service Cycle = (Total Time of Service Cycle – Service Unavailability Duration)/Total Time of Service Cycle x 100%.

Service Availability commitment:

Huawei Cloud will use all commercially and technically reasonable efforts to ensure the Service Availability of an enterprise switch deployed in multiple AZs and in a single AZ in each Service Cycle are, respectively, no less than 99.95% and 99.9%.

An enterprise switch can work together with Direct Connect, Virtual Private Network (VPN), and Virtual Private Cloud (VPC). This SLA only guarantees the service availability of enterprise switches.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the service credits described below:

Enterprise switch deployed in multiple AZs:

Service Availability	Service Credit
99.00% ≤ Service Availability < 99.95%	10% of Monthly Service Fee
95.00% ≤ Service Availability < 99.00%	25% of Monthly Service Fee
Service Availability < 95.00%	100% of Monthly Service Fee

Enterprise switch deployed in a single AZ:

Service Availability	Service Credit
99.00% ≤ Service Availability < 99.9%	10% of Monthly Service Fee
95.00% ≤ Service Availability < 99.00%	25% of Monthly Service Fee
Service Availability < 95.00%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Host Security Service (HSS) Service Specific Terms

1. Definition

Service Unavailability means the HSS system log shows that HSS instances were inaccessible for 5 or more consecutive minutes due to HUAWEI CLOUD.

2. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in Each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the HSS Service Availability Rate in each Service Cycle no less than 99.9%.

3. Service Credit

If we fail to meet the commitment above on HSS Service Availability Rate, HUAWEI CLOUD will provide you with the following service credits:

Service Availability Rate	Service Credit
99% ≤ Service   Availability Rate < 99.9%	15% of Monthly   Service Fee
95% ≤ Service   Availability Rate < 99%	30% of Monthly   Service Fee
Service   Availability Rate < 95%	100% of Monthly   Service Fee

Last Updated: January 15, 2024

IoT Device Access (IoTDA) Service Specific Terms

1. Definitions

Service Unavailability means the IoTDA system log shows that IoTDA was inaccessible for more than five consecutive minutes due to a fault with Huawei Cloud.

Valid Requests refer to the requests received by IoTDA from devices or application servers. Valid requests do not include those failed due to carrier network faults, devices being offline, user misoperations, and specifications limits. For example:

1) After platform APIs are called, non-5XX errors (5XX indicates server system exceptions) are returned due to network disconnection or rate limiting for frequent requests;

2) Devices report data or applications send requests to devices too frequently. The number of requests exceeds the service specifications, for example, 50 queries per second (QPS) when Quality of Service (QoS) 0 is used for a single device;

3) Devices cannot communicate with the cloud due to carrier network faults. For example, devices fail to connect to the cloud and report data, or the cloud fails to send requests to devices;

4) Data is discarded because it is configured incorrectly by users or does not comply with the rule configuration. Access requests fail because the target cloud service or third-party system is abnormal;

5) Requests fail to be processed due to issues of your servers.

Failed Requests refer to the requests failed to be processed due to IoTDA faults. Request failures include device connection errors, device communication errors, and failures to forward or push messages to other cloud services or third-party systems.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Your improper management, misoperations, and negligence, for example:

a.Internal personnel disclose important credentials such as passwords;

b.Users incorrectly modify system parameters or unsubscribe from services using the console or application programming interfaces (APIs);

2) System attacks or device spoofing caused because you do not implement necessary protection according to Huawei's security suggestions or best practices in the industry. For example:

a.Attacks enter the production network from the office network due to inappropriate network isolation;

b.Users use symmetric keys, instead of certificates, for device access;

c.Users use the one-model-one-secret mode with a low security level for device access;

d.Users use unencrypted communication protocols, such as Message Queuing Telemetry Transport (MQTT), Constrained Application Protocol (CoAP), and Hypertext Transfer Protocol (HTTP);

3) Invalid requests;

4) The actual service traffic (number of devices and maximum number of messages per second) exceeds the subscribed service specifications;

5) Major change operations with your authorization.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of IoTDA per Service Cycle be no less than 99.95%

4. Service Credit

If we fail to meet the above commitment, we will provide you with the following service credit.

Service Availability Rate	Coupon Amount
99.9% ≤ Service Availability Rate < 99.95%	5% of Monthly Service Fee
99% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate < 99%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Cloud Firewall (CFW) Service Specific Terms

1. Definition

Service Unavailability: The CFW system log shows that all incoming data packets of a customer's ECS instance are discarded by the CFW due to the faults of Huawei Cloud for five (5) or more consecutive minutes.

2. Disclaimers

For CFW, Service Unavailability in the following conditions cannot be counted in to Service Unavailable Duration, in addition to the conditions described in the definition of Service Unavailability:

1) Attack traffic exceeds the concurrent requests that can be handled by the CFW edition you purchased;

2) Hacker attacks on user applications;

3) Issues caused by a faulty backend server.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of CFW per Service Cycle is not less than 99.95%.

4. Service Credit

If CFW fails to meet the commitment on the service availability rate, Huawei Cloud will compensate you based on the terms given in the following table.

Service Availability	Coupon
99.00% ≤ Service Availability Rate < 99.95%	10% of the basic monthly payment
95.00% ≤ Service Availability Rate < 99.00%	30% of the basic monthly payment
Lower than 95%	100% of the basic monthly fee

Last Updated: January 15, 2024

Cloud Phone Host (CPH) Service Specific Terms

1. Definitions

You understand that Cloud Phone Host (CPH) consists of the following two parts:

• Cloud phone server (paid): includes underlying hardware servers and server system management services.
• Cloud phone (free): includes cloud phones and phone ROM.

Unless otherwise specified, the following SLA applies only to cloud phone servers.

Service Cycle means a calendar month.

Total Time of Service Cycle per Cloud Phone Server means the total minutes in a Service Cycle for a cloud phone server. Total Time of Service Cycle = Total number of days in a Service Cycle x 24 hours x 60 minutes.

CPH Unavailability: The CPH system log shows that CPH was inaccessible for more than five consecutive minutes due to Huawei Cloud reasons.

Unavailable Time per Cloud Phone Server means the total minutes that a cloud phone server is unavailable in a Service Cycle.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Planned system maintenance, such as cutovers, maintenance, upgrades, or fault simulations that Huawei Cloud has notified you of in advance;

2) Huawei Cloud operations authorized by you;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Server hardware faults and your refusal to authorize Huawei Cloud to fix faults;

5) Hardware faults of servers whose warranty has expired and your refusal to upgrade;

6) Your application or installation activities;

7) External devices embedded at the request of you;

8) Server disconnection resulted from normal server operations including but not limited to OS installation, OS reinstallation, BIOS modifications, server shutdowns, or server restarts;

9) Service unavailability that persists when Huawei is waiting for your authorization for server hardware repair and when Huawei engineers are repairing it;

10) A single point of failure (SPOF) on a piece of hardware that does not affect your services due to hardware redundancy. For example, if a single NIC in a bond is faulty or a single disk in a storage redundancy group is faulty, and the faults do not affect your services, the SLA does not apply;

11) Data loss on local disks in cases where, for example, the disks are damaged or the servers break down;

12) Attacks on your applications;

13) Data leakage or loss (including passwords) due to your own actions;

14) Your failure to follow the Huawei Cloud product documentation or other guidelines to perform operations, such as stopping a server, restarting a server, and detaching Elastic Volume Service (EVS) disks from a server, when using APIs, CLI, or console;

15) Service anomalies resulting from your use of third-party services;

16) Cloud phone unavailability due to improper modification of the cloud phone configuration or file (for example, the system service is abnormal due to the modification of the cloud phone ROM).

3. Service Availability

Formula for calculating Service Availability in each Service Cycle

Service Availability per Service Cycle = (Total Time of Service Cycle per Cloud Phone Server – Unavailable Time per Cloud Phone Server)/Total Time of Service Cycle per Cloud Phone Server x 100%.

Service Availability Commitment

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability of a cloud phone server per Service Cycle is not less than 99.5%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the service credit described below:

Service Availability	Service Credit
99% ≤ Service Availability < 99.5%	10% of Monthly Service Fee
Service Availability < 99%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Cloud Eye Service（CES）Service Specific Terms

1. Definitions

Service Unavailability means the Cloud Eye Service (CES) system log shows that a CES was inaccessible for five (5) or more consecutive minutes.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will use commercially and technically reasonable efforts to make the Service Availability Rate of CES per Service Cycle not less than 99.9%

3. Service Credit

If we fail to meet the above commitment on CES Service Availability Rate, Huawei Cloud will provide you with the below service credit:

Service Availability	Service Credit
99%≤Service Availability Rate <99.9%	10% of Monthly Service Fee
95%≤Service Availability Rate <99%	30% of Monthly Service Fee
Service Availability Rate <95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Mix Deposit Service (MDS) Service Specific Terms

1. Definitions

Service unavailability: The customer's servers, switches, or other devices hosted on Huawei Cloud have stopped working due to a failure of facilities in Huawei Cloud equipment rooms.

2. Disclaimers

In addition to the unavailability issues described in general terms, unavailability resulting from the following issues is not included as part of the Service Unavailability Duration:

1) The customer did not deploy dual power supplies and the standalone power supply (including but not limited to power tripping and maintenance) causes aforesaid devices to become unavailable;

2) Devices become unavailable due to operations performed or authorized by the customer or any negligence in operations;

3) The customer's devices were faulty.

3. Service Availability

The service availability in each service cycle is calculated using the following formula:

Service availability in each service cycle = (Total service duration – Service unavailable duration)/Total service duration x 100%.

Service Availability Commitment:

Huawei Cloud will take all commercially viable and technically feasible measures to ensure that the Service Availability of MDS remains at least 99.9% in each service cycle.

4. Service Credit

If Huawei Cloud fails to meet this commitment, we will offer you the following service credits:

Service Availability	Service Credit
99% ≤ Service availability < 99.9%	10% of Monthly Service Fee
Service Availability < 99%	20% of Monthly Service Fee

Last Updated: January 15, 2024

Enterprise Router (ER) Service Specific Terms

1. Definition

Service Unavailability is any time data packets of an enterprise router are discarded for one minute. The Service Unavailable Duration is the number of minutes in a given service cycle that all packets were discarded.

2. Disclaimers

This SLA does not apply to any performance or availability issues:

• caused by network faults of other services (such as the unavailable VPN, Direct Connect, and VPC).
• caused by Direct Connect connections terminated at the same location or a single Direct Connect connection.
• caused by hacker attacks on your applications, including but not limited to DDoS attacks.

3. Service Availability

Service Availability Rate in each Service Cycle is represented by the following formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will use all commercially and technically reasonable efforts to ensure the Service Availability Rate of an enterprise router deployed in multiple AZs and in a single AZ in each Service Cycle is no less than 99.95% and 99.9%.

An enterprise router can be connected to VPCs, Direct Connect, VPN, and Cloud Connect connections. This SLA only guarantees the Service Availability Rate of an enterprise router.

4. Service Credit

If we fail to meet the commitments described above on Enterprise Router Service Availability Rate, Huawei Cloud will provide you with the below service credits:

Enterprise router deployed in multiple AZs:

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
95.00% ≤ Service Availability Rate < 99.00%	25% of Monthly Service Fee
Service Availability Rate < 95.00%	100% of Monthly Service Fee

Enterprise router deployed in a single AZ:

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
95.00% ≤ Service Availability Rate < 99.00%	25% of Monthly Service Fee
Service Availability Rate < 95.00%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Data Replication Service (DRS) Service Specific Terms

1. Definition

Service Unavailability means the DRS system log shows that DRS was inaccessible for five (5) or more consecutive minutes because of Huawei Cloud.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

2) Network faults, device faults, or configuration adjustment of non-Huawei devices;

3) Other force majeure factors.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially viable and technically feasible efforts to ensure that the Service Availability Rate of DRS per Service Cycle is not less than 99.90%.

4. Service Credit

If we fail to meet the commitment above on DRS Service Availability Rate, Huawei Cloud will provide you with the following service credits:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.90%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Cloud Bastion Host (CBH) Service Specific Terms

1. Definitions

Service Unavailability means the CBH system log shows that the CBH service was inaccessible for 5 consecutive minutes or longer because of Huawei Cloud.

2. Disclaimers

1) Customer operations, such as rebooting, deleting, or restoring DB instances through a console or APIs;

2) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Service interruption caused by insufficient CPU and storage resources due to customer's heavy service workload;

5) Hacker attacks on customer applications;

6) Data (including passwords) leakage or loss due to improper maintenance of the customer;

7) Operations authorized by the customer or any mis-judgment in operations;

8) Customer's failure to comply with the product usage documentation or recommendations provided by Huawei Cloud;

9) Service suspension or termination caused by customer's breach of contract (such as an overdue payment);

10) Any force majeure event.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Commitment:

Huawei Cloud will take all commercially viable and technically feasible measures to ensure that the Service Availability of CBH remains at least 99.9% in each service cycle.

4. Service Credit

If we fail to meet these commitments, Huawei Cloud will provide you with the service credit described below:

Service Availability	Service Credit
99% ≤ Service Availability Rate < 99.9%	15% of Monthly Service Fee
Service Availability Rate < 90%	25% of Monthly Service Fee

Last Updated: January 15, 2024

Global Accelerator (GA) Service Specific Terms

1. Service Unavailability

A Global Accelerator is considered unavailable if it fails to establish connections with its endpoints or packets are lost for one or more consecutive minutes.

2. Disclaimers

This SLA does not apply if service unavailability is caused by any abnormality described in the General Terms or either of the following:

• Network failures, device faults, or configuration changes in non-Huawei devices, including but not limited to network carriers' line faults;
• Your applications are attacked by, including but not limited to DDoS attacks;
• Your applications are abnormal.

3. Service Availability

A Global Accelerator's Service Availability refers to the availability of the network between the Huawei Cloud access points and the Huawei Cloud private networks. The network availability between the Huawei Cloud access points and your terminals is guaranteed by the carrier from whom you leased a private line.

The service availability in each service cycle is calculated using the following formula:

Service Availability in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration x 100%.

Huawei Cloud will make all efforts to ensure the service availability of at least 99.95% in each service cycle.

4. Service Credits

If the service availability falls below 99.95% due to reasons caused by Huawei Cloud, Huawei Cloud will offer service credits based on the degree of the unavailability:

Service Availability	Service Credit
99.0% ≤ Service Availability < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability < 99%	25% of Monthly Service Fee
Service Availability < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

FunctionGraph Service Specific Terms

1. Definition

Service Unavailable is a situation where the error rate of FunctionGraph is greater than 5% within five (5) minutes. The Service Unavailable Duration is calculated by multiplying the error rate by five minutes. Error rates less than 5% will be excluded.

Error Rate is calculated by dividing the number of failed requests into the total number of valid requests in five minutes.

Failed Requests are those responded with HTTP status code 500 or 503 (not custom ones in function code) due to internal errors.

2. Disclaimers

In addition to the unavailability issues described in general terms, unavailability resulting from the following issues is not included as part of the Service Unavailable Duration:

1) Requests restricted by flow control due to quota exhaustion;

2) Identity verification or authentication failed, or account in arrears;

3) Non-function request failure;

4) Custom status code 500 or 503 returned;

5) Errors or timeouts caused by users' own function code;

6) Requests failed to reach the server due to non-FunctionGraph reasons.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Cycle - Service Unavailable Duration)/Total Time of Service Cycle x 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of FunctionGraph per Service Cycle is higher than 99.95%.

4. Service Credit

If FunctionGraph fails to meet the preceding commitment, Huawei Cloud will provide you service credit as shown in the following table.

Service Availability	Service Credit
99% ≤ Service Availability Rate ≤ 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: August 22, 2024

You can see what is updated in FunctionGraph Service Level Agreement History Version

Edge Security (EdgeSec) Service Specific Terms

1. Definition

Service Unavailability means the EdgeSec system logs show that, because of Huawei Cloud, the customer service failure rate is higher than 20% and there was a service interruption lasting more than 10 consecutive minutes. An unavailability period of less than 10 minutes is not counted as part of the service unavailability duration.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by:

1) Customer's backend origin servers, for example, bandwidth exhaustion, origin IP address exposure, data center breakdowns, network jitter, installation or upgrades of OSs or applications;

2) Improper design of customer services, for example, if the customer service network is disconnected because their service does not support session synchronization or session persistence;

3) An insufficient EdgeSec protection quota, for example, if the attack traffic exceeds the concurrent requests the EdgeSec edition the customer is using can handle;

4) Non-application layer attacks, for example, heavy-traffic DDoS attacks when the customer did not purchase the EdgeSec Anti-DDoS service;

5) Hacker attacks on customer applications;

6) Consequences of data leakage or loss (including passwords) due to the customer's actions;

7) Service suspension or termination caused by customer's breach of contract, regulations, or laws and other activities, such as security threats or frauds made by the customer or their end users to Huawei Cloud services;

8) An increase in service volume without notice. If the customer's service volume increases sharply (for example, the bandwidth increases or the number of requests increases by 30% or more compared to the previous month), the customer shall notify Huawei Cloud EdgeSec by submitting a service ticket at least three working days in advance.

3. Service Availability

Calculation formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

Huawei Cloud will make all commercially viable and technically feasible efforts to ensure that the Service Availability Rate of EdgeSec per Service Cycle is not less than 99.9%.

4. Service Credit

If EdgeSec fails to meet this commitment to the service availability rate, Huawei Cloud will compensate you based on the terms given in the following table.

Service Availability	Service Credit
99% ≤ Service Availability Rate < 99.9%	10% of monthly service fee
Service Availability < 99%	25% of monthly service fee

Last Updated: January 15, 2024

Video on Demand (VOD) Service Specific Terms

1. Definition

Service Cycle: A service cycle is a calendar month. Failed requests include: (1) Failed HTTP 5XX requests caused by VOD faults; (2) Subscribers' normal requests that fail to reach the VOD server. The following types of requests are not considered failed requests:

1) Playback errors are caused by video file exceptions;

2) Domain names are banned due to content violation or other causes;

3) Playback failed because signed URLs expire when URL validation is enabled;

4) Requests are restricted by VOD due to hacker attacks on applications;

5) Requests are incorrect or services are unavailable caused by reasonable upgrade, change, or suspension initiated by VOD.

Total Time of Service Cycle: The total number of minutes during each service cycle. Total time of service cycle = Total number of days in each service cycle x 24 hours x 60 minutes.

Service Downtime: The total amount of time when the service is unavailable during a service cycle, rounded up to the nearest month. Service downtime spanning multiple service cycles is only counted within the first service cycle.

Service Availability: The percentage of the time when a service is available in a service cycle.

Error rate in an interval of 5 minutes = Number of failed requests in 5 minutes/Total number of requests in 5 minutes x 100%.

Service Unavailability: In an interval of 5 minutes, there are more than 100 requests and the error rate calculated based on the data recorded in VOD logs is greater than 0.05% due to a fault with Huawei Cloud.

2. Disclaimers

VOD does not limit the traffic, bandwidth, transcoded output duration, or storage. You can expand the capacity based on service requirements. If the service volume is expected to increase sharply (as shown in the following metrics), you need to notify Huawei Cloud VOD by submitting a service ticket at least three working days in advance. Otherwise, the availability may be affected and no compensation can be obtained.

Metrics related to sudden service volume increase:

Bandwidth: The peak growth is expected to be greater than 10 Gbit/s, and the growth is expected to be concentrated in some regions or carriers.

Storage: The storage growth is expected to be greater than 10 TB.

Transcoding: The transcoded output duration is expected to increase by more than 10,000 minutes per day.

3. Service Availability

Service availability rate in each service cycle is represented by the following formula:

Service availability rate in each service cycle = (Total time of service cycle – Service downtime)/Total time of service cycle x 100%.

Service availability rate commitment:

Huawei Cloud will make all commercially and technically reasonable efforts to ensure a VOD service availability rate of at least 99.9% per service cycle.

4. Service Credit

If VOD fails to fulfill the preceding commitment, Huawei Cloud will provide you with service credit as shown in the following table.

Service Availability Rate	Service Credit
99% ≤ Service availability rate < 99.9%	10% of Monthly service fee
Service availability rate < 99%	25% of Monthly service fee

Last Updated: January 15, 2024

Web3 Node Engine Service (NES) Service Specific Terms

1. Definition

Service Cycle is the total time for a user to use a node instance.

Total Time of Service Cycle per NES node is the total number of minutes in a Service Cycle for an NES node. Total Time of Service Cycle = Total number of days in a Service Cycle x 24 hours x 60 minutes.

Service Unavailability means that the NES system log shows that NES was inaccessible for more than ten consecutive minutes due to a fault with Huawei Cloud.

Monthly Service Fee means service fees paid by customers for an NES node in a Service Cycle.

2. Disclaimers

1) Service Unavailability does not include unavailability resulting from user clients, smart contracts, validators, or MEV-Boost that is not caused by Huawei Cloud;

2) Service Unavailability does not include unavailability resulting from misoperation and constraints of node resources.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Cycle – Service Unavailable Duration)/Total Time of Service Cycle x 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of NES per Service Cycle is not less than 99.95%.

4. Service Credit

If we fail to meet the commitments described here for the NES service availability rate, Huawei Cloud will provide you with the service credit below.

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
Service Availability Rate < 99%	25% of Monthly Service Fee

Last Updated: January 15, 2024

KooMessage Service Specific Terms

1. Definition

Service Cycle: a natural month.

Monthly Service Fee: Service fees paid by customers for KooMessage in a Service Cycle.

Service Unavailability: According to the log records in the KooMessage system. KooMessage service cannot be accessed for more than 5 minutes due to HUAWEI CLOUD and any unavailability period for less than 5 minutes is not included.

2. Disclaimers

Service Unavailability caused by the following situations shall not be counted in the Service Unavailability:

1) You fail to follow the KooMessage documentation or refuse to follow the suggestions after being informed of the risks;

2) You violate applicable laws and regulations, requirements of relevant authorities, or rules of the Huawei Cloud platform;

3) Data, passwords, or other information getting lost or leaked due to improper maintenance or handling of confidential information;

4) Overload of client business, requests are flow controlled as requests exceed the specified quota;

5) System maintenance with prior notice from Huawei Cloud to clients, including but not limited to cutover, maintenance, upgrade, patch and simulated troubleshooting;

6) Your message is regarded as a nuisance by the terminal receiving customer.

3. Service Availability

Formula for calculating Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration * 100%.

Commitment for Service Availability:

Huawei Cloud will ensure the service availability rate in each service cycle is at least 99.95% through technically and commercially reasonable efforts.

4. Service Credit

If we fail to meet the above commitment, we will provide you the service credit in the following table: 

Service Availability Rate	Service Credit
95.00% ≤ Service Availability < 99.95%	10% of monthly service fee
Service Availability < 95.00%	25% of monthly service fee

Last Updated: January 15, 2024

Workspace Service Specific Terms

1. Definition

Service Unavailability: The Workspace service is considered unavailable when the logs in the Workspace service system show that all Workspace connection events fail for more than three consecutive minutes due to Huawei Cloud's reasons. The Service Unavailable Duration less than three minutes is not counted.

2. Disclaimers

For Workspace, in addition to the conditions described in the definition of Service Unavailability, Service Unavailability in the following conditions cannot be counted in to Service Unavailable Duration:

1) Unavailability caused by your own operations using APIs or on the console or login page, such as stopping, restarting, and rebuilding a desktop, detaching disks, disabling NICs, and disabling processes, software, and files related to Workspace;

2) Unavailability caused by software or drivers installed by you, third-party software or drivers that are not directly managed by Huawei Cloud, or configuration errors;

3) Unavailability caused by network faults, such as unreachable Internet, unreachable VPN or Direct Connect connection, incorrect VPC security group, firewall policy, or ACL settings, and DNS resolution failures;

4) Unavailability caused by faults of your authentication servers, such as issues of AD or other authentication services, DNS, or certificates;

5) Unavailability caused by OS, such as OS patch installation, OS upgrade, and blue screen of death (BSOD);

6) Unavailability caused by operations authorized by the customer or any misjudgment in operations;

7) After a pay-per-use cloud desktop is shut down, the vCPUs and memory of the cloud desktop will be released. When the cloud desktop is restarted, the restart will fail due to insufficient resources.

3. Service Availability

Calculation formula for the Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will take all commercially viable and technically feasible measures to ensure that the Service Availability Rate of Workspace remains at least 99.9% in each service cycle.

4. Service Credit

If we fail to meet these commitments, Huawei Cloud will provide you with the service credit described below:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.9%	10% of Monthly Service Fee
Service Availability Rate < 99%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Multi-Site High Availability Service (MAS) Service Specific Terms

1. Definition

Service Unavailable means the MAS system log shows that MAS was inaccessible for more than five consecutive minutes due to a fault with Huawei Cloud.

2. Disclaimers

Service Unavailable does not include unavailability resulting from the following issues:

1) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

2) Network faults, device faults, or configuration adjustment of non-Huawei Cloud devices;

3) Hacker attacks on your applications;

4) Customer's failure to comply with the product usage documentation or recommendations provided by Huawei Cloud;

5) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

6) Data (including passwords) leakage or loss due to improper maintenance or storage of the customer;

7) Operations authorized by the customer or any mis-judgment in operations;

8) Any force majeure event.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate Commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of MAS per Service Cycle is not less than 99.5%.

4. Service Credit

If we fail to meet the commitment described here for the MAS Service Availability Rate, Huawei Cloud will provide you with the service credit below.

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.5%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	30% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: January 15, 2024

Global SIM Link (GSL) Service Specific Terms

1. Definitions

Service Unavailability means the GSL system logs show that GSL was inaccessible for more than five consecutive minutes due to a fault with Huawei Cloud.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by:

1) Planned system maintenance, such as cutovers, maintenance, upgrades, or fault simulations that Huawei Cloud has notified you of in advance;

2) Network faults, device faults, or configuration adjustment of non-Huawei devices;

3) Hacker attacks on your applications;

4) Data (including passwords) leakage or loss due to your own actions;

5) Your violations of the information control, traffic control, or information security control rules of the platform;

6) Your failure to comply with the product documentation or usage recommendations of this Service;

7) Content that contains illegal information elements such as pornography, gambling, drugs, politically sensitive information, and fraud;

8) Control measures taken by regulators or carriers;

9) Major events or promotions that Huawei Cloud has notified you of in advance;

10) SIM card suspension due to your violation of laws and regulations, government policies, or infringement of third-party rights and interests;

11) Other force majeure factors.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of Global SIM Link is not less than 99.5%.

4. Service Credit

If we fail to meet the commitment above for the GSL Service Availability Rate, Huawei Cloud will compensate you with the service credits described below:

Service Availability Rate	Service Credit
95% ≤ Service Availability Rate < 99.5%	10% of Monthly Service Fee
Service Availability Rate < 95%	20% of Monthly Service Fee

Last Updated: January 15, 2024

Message & SMS Service (MSGSMS) Service Specific Terms

1. Definition

Service Cycle: If you apply for the Service for the first time and establish an SMS application, the first Service Cycle starts from the date when the application is enabled and ends at the end of the corresponding month. Subsequently, each calendar month is a Service Cycle.

Monthly Service Fee: Total service fees you paid for the Service in a Service Cycle.

Number of Valid SMSs: Number of the split SMSs that are successfully sent to Valid Numbers. If an SMS exceeds the length limit, the Huawei Cloud platform splits the SMS into multiple short messages.

Clear Failure: The valid numbers you submitted were not delivered to the downstream carriers or corresponding service providers due to a Huawei Cloud platform problem.

Valid Number: Numbers you submit that can be used, excluding:

1) Numbers whose carrier IDs are abnormal, such as empty numbers and suspended numbers;

2) Terminal numbers that cannot receive SMSs due to user behavior or other reasons, including numbers that are in arrears or out of service, belong to switched-off phones or phones with full storage space, do not subscribe to or unsubscribe from the SMS service, are with unstable network signals, and are blocked from receiving calls or messages;

3) Blocklisted numbers of Message & SMS or carriers.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Your faults or other reasons;

2) Hackers or malicious attacks on your applications or platforms;

3) Data (including passwords and keys) leakage/loss due to your improper maintenance. Related consequences include SMS bombing;

4) Your failure to comply with the product usage documentation or recommendations provided by Huawei Cloud Message & SMS;

5) Your negligence or operations authorized by you;

6) Your violations against the platform rules of the Service;

7) Your failure to notify Huawei Cloud of your major service changes or high-concurrency requirements in advance;

8) Your service that is regarded as harassment by mobile phone users;

9) Your use of the Service free of charge in some ways, such as using test coupons;

10) Control or requirements of communication management organizations and carriers in the destination country or region;

11) Illegal content or serious complaints;

12) Events caused by major accidents or faults of relevant Internet service providers;

13) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

14) Major events or promotions that Huawei Cloud has notified you of in advance;

15) Non-Huawei Cloud network failures, device faults, or configuration changes, including device, software, and/or technical issues caused by you or any third party (not directly managed by Huawei);

16) Other non-Huawei Cloud issues;

17) Situations that Huawei Cloud can be exempted from liability or compensation as described in relevant laws and regulations, agreements, rules, or rules and descriptions released by Huawei Cloud;

18) Force majeure, accidents or events beyond our reasonable control, including earthquakes, epidemics, interruption of submarine communication cables, telecom infrastructure or system faults, and natural disasters.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Number of Valid SMSs in the Service Cycle – Number of Clear Failures in the Service Cycle)/Number of Valid SMSs in the Service Cycle x 100%.

Service Availability commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of Message & SMS per Service Cycle is not less than 95%. In the event of specific agreements, the agreed Service Availability Rate shall prevail.

4. Service Credit

If we fail to meet these commitments, Huawei Cloud will provide you with the service credit described below:

Service Availability Rate	Service Credit
90% ≤ Service Availability Rate < 95%	10% of Monthly Service Fee
Service Availability Rate < 90%	30% of Monthly Service Fee

Last Updated: January 15, 2024

Global EIP Service (Gip) Service Specific Terms

1. Definition

Service Unavailability means the global Elastic IP (EIP) system log showing that all data packets in the outbound direction are discarded at the Huawei Cloud egress gateway for one (1) or more consecutive minutes due to the fault of Huawei Cloud.

Monthly Expenditure is your total expenditures for a single global EIP and its bandwidth within a Service Cycle (a calendar month). If you pay for several Service Cycles at a time, the Monthly Expenditure is calculated by dividing the total amount billed by the number of months the global EIP and its bandwidth were used.

2. Disclaimers

Service Unavailability does not include unavailability resulting from any of the following:

1) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations;

2) Non-Huawei Cloud network failures, device faults, or configuration changes, including but not limited to service unavailability caused by network carrier line faults;

3) Application attacks including but not limited to DDoS attacks, SYN flood attacks, and virus attacks, or software and hardware compatibility issues;

4) Your services exceeding the supported resource specifications;

5) Faulty backend servers;

6) Loss or disclosure of data, including passwords, due to your improper maintenance;

7) Operations authorized by you or any mis-judgment in operations;

8) Your failure to follow to the product usage documentation or recommendations provided by Huawei Cloud;

9) Force majeure or other accidents that are unforeseeable, insurmountable, and unavoidable and have significant impact on one party or both parties, including but not limited to natural disasters (such as floods, earthquakes, and plague epidemics) and social events (such as wars, unrest, government actions, interruption of telecom backbone lines, hackers, network congestion, technical change of telecommunications departments, and government policies);

10) Line faults due to any reason where Huawei Cloud has notified you that cross-region and cross-line disaster recovery is required, but you did not implement disaster recovery.

3. Service Availability

Service Availability Rate in each Service Cycle is represented by the following formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate Commitment:

Huawei Cloud will use commercially and technically reasonable efforts to ensure the Service Availability Rate of dynamic BGP global EIPs, static BGP global EIPs, and global EIPs of China Telecom, China Unicom, or China Mobile per Service Cycle are, respectively, no less than 99.95%, 99%, and 99%.

4. Service Credit

If we fail to meet the above commitment, Huawei Cloud will provide you with the below service credit:

Global EIPs of the dynamic BGP type:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Global EIPs of the static BGP type:

Service Availability Rate	Service Credit
98% ≤ Service Availability Rate < 99%	10% of Monthly Service Fee
Service Availability Rate < 98%	25% of Monthly Service Fee

Global EIPs of China Telecom, China Unicom, or China Mobile:

Service Availability Rate	Service Credit
98% ≤ Service Availability Rate < 99%	10% of Monthly Service Fee
Service Availability Rate < 98%	25% of Monthly Service Fee

Last Updated: January 15, 2024

SecMaster Service Specific Terms

1. Definition

Service Unavailability means the SecMaster system log shows that SecMaster was inaccessible for five (5) or more consecutive minutes because of Huawei Cloud.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) User operations, such as restarting, deleting, and restoring SecMaster instances through the console or APIs, and building custom models, playbooks, data access, plug-ins, large screens, reports, and page layouts in SecMaster;

2) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei Cloud devices;

4) Hacker attacks on user applications;

5) Data (including passwords) leakage or loss due to improper maintenance of the customer;

6) Operations authorized by the customer or any misjudgment in operations;

7) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

8) Unavailability caused by known problems or risks that have been reported to customers for rectification but have not been resolved;

9) Unavailability caused by external environment, for example, restart or jitter of underlying resources.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate Commitment:

Huawei Cloud shall make commercially and technically reasonable efforts to make the Service Availability Rate for SecMaster up to 99.9%. Should Huawei Cloud fail to ensure the promised rate, Huawei Cloud will compensate you in accordance with Clause 3 of this Agreement.

4. Service Credit

If SecMaster fails to meet this commitment to the service availability rate, Huawei Cloud will compensate you based on the terms given in the following table.

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.9%	10% of monthly service fee
Service Availability Rate < 99%	25% of monthly service fee

Last Updated: January 15, 2024

ServiceStage Service Specific Terms

1. Definition

Service Unavailable

If all consecutive attempts to connect to a specified ServiceStage instance or read/write requests fail within a minute and the failure lasts for more than 1 minute, the ServiceStage instance is unavailable within the minute. The unavailability duration less than 10 minutes is not counted.

2. Disclaimers

In addition to the unavailability issues described in general terms, unavailability resulting from the following issues is not included as part of the Service Unavailable Duration:

1) the region cannot form 3AZ, and an AZ-level fault occurs;

2) caused by loss or disclosure of data, including passwords, due to your improper maintenance;

3) caused by your negligence or allowed operations;

4) instances exceed specifications allowed;

5) caused by third-party open-source software.

3. Service Availability

Formula for Service Availability Percentage in each Service Cycle:

Service Availability Percentage in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Percentage Commitment:

Huawei Cloud will use all commercially and technically reasonable efforts to ensure that the Service Availability Percentage of ServiceStage per Service Cycle is not lower than 99.95%.

4. Service Credit

If we fail to meet the commitments described here for the ServiceStage service availability percentage, Huawei Cloud will provide you with the service credit below.

ServiceStage

Service Availability	Service Credit
99.00% ≤ Service Availability < 99.95%	15% of Monthly Service Fee
95.00% ≤ Service Availability < 99.00%	30% of Monthly Service Fee
Service Availability < 95.00%	100% of Monthly Service Fee

Last Updated: January 15, 2024

CodeArts PerfTest Service Specific Terms

1. Definition

Service only refers to the services described in this SLA.

Service Cycle means a calendar month.

Total Time of Service Cycle means the total minutes during every Service Cycle.

Service Unavailable will be defined for each Service in the Service Specific Terms thereafter. Service Unavailable does not include the situation that Cloud Service is unavailable due to the disclaimers described below and disclaimers in the following service specific terms.

Service Unavailable Duration means the total amount of time when the Service is unavailable during a Service Cycle, rounded up to the nearest month. Service downtime spanning multiple service cycles is only counted within the first service cycle.

Service Availability means during a Service Cycle, the service available percentage per Service Cycle achieved by the Service given in this SLA.

Monthly Service Fee means the service fees you paid for a Service in a Service Cycle. For example, if you pay the service fee of several months at a time, the Monthly Service Fee is calculated as the total service fees paid divided by the total number of months.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Incorrect configuration or pressure test script by you, which hinders the pressure test;

2) Planned system maintenance, such as cutover, maintenance, upgrade, and fault simulations, of which Huawei Cloud has notified the customer in advance;

3) Network faults, device faults, or configuration adjustment of non-Huawei devices;

4) Hacker attacks on your applications or data;

5) Service suspension or termination due to your breach of service clauses (such as arrears);

6) Data (including passwords) leakage or loss due to your improper maintenance or storage;

7) Operations authorized by you or any mis-judgment in operations.

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate Commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of CodeArts PerfTest per Service Cycle is not less than 99.5%.

4. Service Credit

If we fail to meet these commitments, Huawei Cloud will provide you with the service credit described below:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.5%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	30% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: March 28, 2024

Video Intelligent Analysis Service (VIAS) Service Specific Terms

1. Definition

Service Unavailability means the VIAS system log shows that VIAS was inaccessible for five (5) or more consecutive minutes due to a fault with Huawei Cloud. Duration of the inaccessibility and incapability that lasts shorter than five (5) minutes shall not be included.

2.Disclaimers

Service Unavailability caused by the following situations shall not be counted in the Service Unavailability Duration:

1) The Service Unavailability is caused by planned system maintenance, including but not limited to cutover, maintenance, upgrade, patching, and fault simulations, of which Huawei Cloud has notified you in advance;

2) The Service Unavailability is caused by hacker attacks on your applications;

3) Data or password leakage or loss is caused due to your improper maintenance or improper use of confidential information;

4) The Service is interrupted or unavailable due to the OS upgrade, Docker software upgrade, or IEF edge software upgrade of your edge node;

5) The Service is unavailable due to non-Huawei factors, such as network faults (for example, edge node network disconnection), hardware faults (for example, hard disk damage), OS faults, Docker software faults, insufficient Docker storage space, and mismatch of inference card drivers;

6) The Service is unavailable because the camera installation, video stream encoding format, video stream artifacts, video stream quality, video image blocking, or other items do not comply with algorithm constraints (including but not limited to https://support.huaweicloud.com/intl/en-us/productdesc-vias/vias_01_0007.html);

7) The tasks failed or cannot be deleted due to network or server restart.

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%.

Service Availability Rate Commitment:

Huawei Cloud will take all commercially viable and technically reasonable efforts to make the Service Availability Rate of VIAS per Service Cycle not less than99.9%.

4. Service Credit

If we fail to meet the above commitment on VIAS Service Availability Rate, Huawei Cloud will provide you with the below Service Credit.

Service Availability Rate	Service Credit
99.00% ≤ Service Availability Rate < 99.90%	10% of Monthly Service Fee
Service Availability Rate < 99.00%	25% of Monthly Service Fee

Last Updated: May 22, 2024

KooDrive Service Specific Terms

1. Definition

Service Unavailable means the KooDrive log shows that KooDrive was inaccessible for more than five minutes due to a fault with Huawei Cloud.

2. Disclaimers

This SLA also does not apply to any performance or availability issues caused by the following:

(1) User misoperations or network constraints;

(2) A force majeure

3. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

Huawei Cloud will make commercially and technically reasonable efforts to ensure that the Service Availability Rate of KooDrive per Service Cycle is not less than 99.95%.

4. Service Credit

If we fail to meet the commitment above on KooDrive Service Availability Rate when you subscribe to this service independently, Huawei Cloud will provide you service credits as shown in the following table:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
Service Availability Rate < 99%	25% of Monthly Service Fee

Last Updated: August 1, 2024

Scalable File Service Turbo  (SFS Turbo) Service Specific Terms

1. Definitions

Service Unavailability means the SFS Turbo system log shows that SFS Turbo was inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD.

2. Service Availability

Service Availability Rate in each Service Cycle calculation formula:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration) / Total Time of Service Duration * 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of SFS Turbo per Service Cycle not less than 99.95%.

3. Service Compensation

If we fail to meet the above commitment, we will provide you the service credit in the following table:

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	30% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: August 22, 2024

 

KooPhone Service Specific Terms

1. Definition

Service Unavailable means the KooPhone system log shows that KooPhone was inaccessible for five consecutive minutes or more due to a fault with Huawei Cloud.

2. Disclaimers

In addition to the unavailability issues described in general terms, unavailability resulting from the following issues is not included as part of the Service Unavailable Duration:

1) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations

2) Network faults, device faults, or configuration adjustment of non-Huawei Cloud devices

3) Hacker attacks on your applications

4) Data (including passwords) leakage or loss due to your improper maintenance or storage

5) Operations authorized by you or any mis-judgment in operations

6) Your failure to follow to the product usage documentation or recommendations provided by Huawei Cloud

7) Service suspension or termination due to the customer's breach of service clauses (such as arrears)

8) Service function or performance exceptions due to the exceeded quota limit of KooPhone

9) Any force majeure event

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%

Service Availability Rate Commitment:

Huawei Cloud will make all commercially viable and technically feasible efforts to ensure that the Service Availability Rate of KooPhone per Service Cycle is not less than 99.95%.

4. Service Credit

If KooPhone fails to meet the preceding commitment, Huawei Cloud will provide you service credit as shown in the following table.

Service Availability	Service Credit
99% ≤ Service Availability Rate < 99.95%	3% of Monthly Service Fee
Service Availability Rate < 99%	5% of Monthly Service Fee

Last updated: August 22, 2024

 

Cloud Application Engine (CAE) Service Terms

1. Definition

Service Unavailable means that the components deployed on CAE are unavailable for more than 10 consecutive minutes due to Huawei Cloud reasons according to the logs in the CAE service system. Any unavailable duration less than 10 minutes is not counted. Or else, all data packets outbound from CAE are discarded by the Huawei Cloud egress gateway for more than one minute due to Huawei Cloud reasons. Unavailable duration less than one minute is not counted.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

1) Planned system maintenance that Huawei Cloud has notified you of in advance, such as cutovers, maintenance, upgrades, or fault simulations

2) Non-Huawei Cloud network failures, device faults, or configuration changes, including but not limited to service unavailability caused by network carrier line faults

3) Caused by hacker attacks on your applications, including but not limited to DDoS or SYN Flood attacks

4) Data (including passwords) leakage or loss due to your improper maintenance or storage

5) Operations authorized by you or any mis-judgment in operations

3. Service Availability

Formula for Service Availability Rate in each Service Cycle:

Service Availability Rate in each Service Cycle = (Total Time of Service Duration – Service Unavailable Duration) / Total Time of Service Duration x 100%

Service Availability Rate Commitment:

Huawei Cloud will use commercially and technically reasonable efforts to ensure the service availability rate is at least 99.8% for CAE in each service cycle.

4. Service Credit

If we fail to meet the commitments described here for the CAE service availability rate, Huawei Cloud will provide you with the service credit below.

Service Availability	Service Credit
99% ≤ Service Availability Rate < 99.8%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

 

Last updated: November 21, 2024

 

Cloud Container Instance (CCI) Specific Terms

1. Definitions

Service Unavailability of a Single Pod: System logs show that the workload is inaccessible for one (1) or more consecutive minutes because of Huawei Cloud. Service Unavailability does not include unavailability or service suspension caused by system maintenance (including cutovers, maintenance, and upgrades), or unavailability caused by you or a third party, or by force majeure.

Service Unavailability of Pods in Multiple AZs of the Same Region: System logs show that the workloads are inaccessible for one (1) or more consecutive minutes because of Huawei Cloud. Service Unavailability does not include unavailability or service suspension caused by system maintenance (including cutovers, maintenance, and upgrades), or unavailability caused by you or a third party, or by force majeure. When you use CCI to run workloads in multiple AZs of the same region, if the workloads in all AZs are unavailable at the same time, all pods in that region are considered unavailable.

2. Disclaimers

In addition to the situations specified in General Terms, the following situations shall not be counted in the Service Downtime:

1) Workloads are unavailable due to your container images, software or drivers you install, third-party software or drivers that are not directly managed by Huawei Cloud, configuration errors, or hacker attacks.

2) Workloads are unavailable because the data in the temporary storage media described in the CCI documentation is lost (for example, when the local disk is damaged).

3) Workloads are unavailable because resources are sold out, which CCI will display an event for.

4) Workloads are unavailable because the quotas and capacities of CCI and dependent services (such as network and storage services) exceed the limits in your plan.

5) CCI service availability is undermined due to unavailability of its dependent services, including compute, storage, and networking services. (If this happens, refer to the SLAs of the dependent services.)

6) Workloads are unavailable due to component exceptions in your on-premises data center when workloads in your on-premises data center are scaled to CCI through the add-on.

7) Workloads are unavailable because data, passwords, or other authentication credentials are lost or leaked due to improper maintenance or confidentiality measures.

8) CCI is abnormal or fails to meet your requirements because you change the configurations without following the CCI documentation or console instructions, unless otherwise specified. There are no SLAs for the open beta capabilities specified in the CCI documentation or on the console.

3. Service Availability

Formula for Service Availability in each Service Cycle:

Service Availability per Service Cycle = (Total Time of Service Cycle – Service Downtime)/Total Time of Service Cycle × 100%

Service Availability Commitment:

Huawei Cloud shall make reasonable commercial and technical efforts to promise that the service availability of a single pod per Service Cycle is not lower than 99.95% and that of pods in multiple AZs of the same region per Service Cycle is not lower than 99.99%.

4. Service Credit

If we fail to meet the commitment, Huawei Cloud will provide you with the service credits described below.

A single pod:

Service Availability	Service Credit
99% ≤ Service Availability < 99.95%	10% of Monthly Service Fee
95% ≤ Service Availability < 99%	25% of Monthly Service Fee
Service Availability < 95%	100% of Monthly Service Fee

Pods in multiple AZs of the same region:

Service Availability	Service Credit
99% ≤ Service Availability < 99.99%	10% of Monthly Service Fee
95% ≤ Service Availability < 99%	25% of Monthly Service Fee
Service Availability < 95%	100% of Monthly Service Fee

Last updated: November 25, 2024

 

TaurusDB Service Specific Terms

1. Definition

Service Unavailability means the TaurusDB system log shows that a DB instance was inaccessible for one (1) or more consecutive minutes because of HUAWEI CLOUD. Duration of the inaccessibility and inability that lasts for shorter than one (1) minute shall not be included.

2. Disclaimers

Service unavailability caused by the following situations shall not be counted in the service unavailability period:

1) Customer operations, such as rebooting, deleting, or restoring TaurusDB instances through the console or APIs;

2) Network faults, device faults, or configuration adjustment of non-Huawei devices;

3) Faults that occur due to single-AZ deployment, instead of multi-AZ deployment;

4) Hacker attacks on customer applications;

5) Planned system maintenance, such as cutover, maintenance, upgrade, and fault drills, of which Huawei Cloud has notified the customer in advance;

6) Data (including passwords) leakage or loss due to improper maintenance of the customer;

7) Operations authorized by the customer or any misjudgment in operations;

8) Customer failure to comply with TaurusDB documentation or recommendations;

9) Service interruptions caused by insufficient vCPUs, memory, storage, and maximum number of connections of a DB instance or proxy instance due to heavy workloads of the customer;

10) Faults caused by open-source community bugs or other non-Huawei Cloud issues;

11) Service suspension or termination due to the customer's breach of service clauses (such as arrears);

12) Customer failure to rectify known problems or risks which Huawei Cloud has notified them of;

13) Other force majeure factors.

3. Service Availability

Service Availability Rate in Each Service Cycle calculation formula:

Service Availability Rate in Each Service Cycle = (Total Time of Service Duration - Service Unavailable Duration)/Total Time of Service Duration x 100%.

Service Availability Rate commitment:

HUAWEI CLOUD will use commercially and technically reasonable efforts to make the Service Availability Rate of TaurusDB per Service Cycle be no less than 99.99%.

Remarks: HUAWEI CLOUD cannot ensure the high availability of a single-node DB instance (with a primary node or a HTAP read replica) because the instance is deployed on only a single server.

4. Service Credit

If we fail to meet the commitment above on TaurusDB Service Availability Rate, HUAWEI CLOUD will provide you with service credit as follows.

Service Availability Rate	Service Credit
99% ≤ Service Availability Rate < 99.99%	10% of Monthly Service Fee
95% ≤ Service Availability Rate < 99%	25% of Monthly Service Fee
Service Availability Rate < 95%	100% of Monthly Service Fee

Last Updated: November 30, 2024

You can see what is updated in TaurusDB Service Level Agreement History Version

CloudPond Service Specific Terms

1. Definition

Service definition: CloudPond is a distributed cloud solution provided by Huawei Cloud in customer equipment rooms.

Service availability scope: Huawei Cloud services that run on CloudPond.

Definition of service unavailability: As CloudPond is a part of Huawei Cloud, Huawei Cloud performs unified O&M. After eliminating the restrictions that affect CloudPond unavailability situations, services running on CloudPond will comply with the definition of unavailability of services running on Huawei Cloud.

If other cloud services have special conventions for the CloudPond scenario, the special conventions prevail.

2. Disclaimers

This SLA does not apply to any performance or availability issues caused by the following:

(1) The CloudPond cloud-edge network cannot communicate properly due to user-side faults. This is not considered as a service unavailable.

Common scenarios include but are not limited to the following:

a. Communication failure occurs on the network provided by the carrier used by the user.

b. Communication faults caused by incorrect configuration of data center network devices.

c. Communication faults caused by malicious attacks on the user network;

d. Communication faults occur due to network arrears or frequent changes of public IP addresses.

(2) If the hardware of the CloudPond cabinet is faulty due to user-side problems, the service is not unavailable.

Common scenarios include but are not limited to the following:

a. Users open CloudPond cabinets and adjust, move, or damage hardware without permission.

b. Users move CloudPond cabinets ithout prior communication with the Huawei Cloud team.

c. Users change cables (such as power supply and network cables) of the CloudPond cabinet without prior communication with the Huawei Cloud team.

d. Users power off the entire cabinet without prior communication with the Huawei Cloud team.

(3) If the equipment room where the CloudPond cabinet is located does not meet the CloudPond normal running standard due to user-side problems, this situation is not considered as a service unavailable.

Common scenarios include but are not limited to the following:

a. The power supply in the customer's equipment room is unstable and does not meet the CloudPond operation requirements.

b. The cooling system in the customer equipment room is faulty and cannot provide temperature and humidity suitable for the normal operation of the CloudPond.

c. Vulnerabilities exist in the management work of the customer's equipment room. As a result, the CloudPond cabinet is maliciously damaged.

(4) No standby server is configured for the CloudPond computing resources used by the user. As a result, the CloudPond cannot provide services when the related hardware is faulty, and the service is not unavailable.

(5) CloudPond cannot provide services due to security attacks because users do not complete security protection on the application/self-workload side. This situation is not considered as a service unavailable situation.

Common scenarios include but are not limited to the following:

a. Security hardening is not performed on ECS VMs running on CloudPond. As a result, viruses and Trojan horses are attacked and the OS breaks down.

b. Users do not correctly configure security services, VPC networks, and port rules. As a result, user services are maliciously attacked and become unavailable.

c. The user's service information is stolen because the user uses simple passwords or the password and AK/SK are not properly stored.

(6) CloudPond cannot provide services because users do not pay for or renew resources in a timely manner. Therefore, CloudPond cannot provide services.

(7) CloudPond cannot provide services due to technical problems during service deployment, upgrade, change, backup and restoration, and DR drills.

(8) If the CloudPond cannot provide services before capacity expansion and acceptance are complete due to capacity expansion requirements, the CloudPond cannot provide services.

(9) The CloudPond cloud-edge network connection solution needs to be switched due to Huawei Cloud or customer reasons. During the switchover, CloudPond cannot provide services properly. Therefore, the service is not unavailable.

(10) If the CloudPond service cannot run properly due to user operation reasons, the CloudPond cannot provide services during the working period, such as configuration restoration, service startup, and data verification, after the user-side fault is rectified, the CloudPond service cannot be regarded as unavailable.

(11) Due to Huawei Cloud version upgrade requirements, CloudPond cannot provide services during the upgrade period agreed with the customer. Therefore, the service is not unavailable.

(12) In the CloudPond scenario where the customer leases equipment rooms (including Huawei equipment rooms), network unavailability caused by carriers should not be counted in the service unavailability time.

(13) If the CloudPond is deployed in the equipment room leased by the customer (including the equipment room leased by Huawei), service unavailability caused by network faults, device configuration adjustment, or mains power interruptions other than Huawei Cloud equipment (including the equipment leased by Huawei) shall not be counted in the service unavailability time.

(14) If CloudPond cannot provide services due to other non-Huawei Cloud reasons, the service unavailability shall not be counted.

3. Service availability

CloudPond is a part of the Huawei Cloud and is operated and maintained by Huawei Cloud Operation Team in a unified manner. After the disclaimers above that affect the availability of CloudPond are excluded, the availability of Huawei Cloud services running on CloudPond is ensured, and it complies with the availability standards of Huawei Cloud cloud services in the single Availability Zone (AZ) model.

If other cloud services have special conventions for the CloudPond scenario, the special conventions prevail.

4. Compensation for services

CloudPond is a part of the Huawei Cloud and is operated and maintained by Huawei Cloud Operation Team in a unified manner. If the Huawei Cloud services running on CloudPond do not meet the service commitments, Compensation will be made based on the service compensation standard of each Huawei Cloud service.

If other cloud services have special conventions for the CloudPond scenario, the special conventions prevail.

 

Last Updated: December 3, 2024