Zilliz Cloud SLA
Zilliz, Inc. ("we", "us", or "our") will use reasonable efforts to ensure the maximum availability of Zilliz Cloud Dedicated (Enterprise)("Platform"), which is one of the cloud services we offer. This Service Level Agreement ( this "SLA") is an integral part of the Terms of Service (the "Agreement") and applies only to the extent expressly set forth in the Agreement. Any capitalized terms used in this SLA are as defined in the Agreement unless otherwise specified. Zilliz Cloud may modify this SLA from time to time.
1. Definition
Month: A month that divides a year into twelve equal parts, which is 30.41 days;
Monthly Minutes: The total number of minutes in a month;
Operation Window Minutes: The number of minutes of planned downtime for operation and upgrades of the Zilliz Cloud cluster in a month;
Service Downtime: The service Downtime due to server failure outside the normal operation window. That is, if all consecutive attempts by Zilliz Cloud's monitoring system to connect to a cluster fail within one minute, or if there are server query or write request failures due to server failures (non-current limiting factors or user-side error calls) within one minute, then the cluster is considered unavailable for the minute.
Monthly Service Fee: The total amount of service fees you pay in a month;
2. Response Time
| Problem Level | Problem Definition | Dedicated(Enterprise) Response Time |
|---|---|---|
| Urgent | These are critical problems that require immediate attention, as they significantly impact the functionality or availability of the service. Examples include:
| 1 hour |
| High | High-priority issues are those that cause service quality degradation. Examples include:
| 4 hours |
| Normal | Normal-priority issues usually involve non-urgent matters related to the functionality or user experience of the service. Examples include:
| 1 Business day |
3. Service Availability
3.1 Calculation of Service Availability
Service availability refers to the monthly minutes minus the operation window minutes, divided by monthly minutes. Calculated as:
Service Availability = (Monthly Minutes - Operation Window Minutes) / Monthly Minutes
When the database is deployed only part of the time in a month, it is presumed that the operation window minutes of that month are distributed in the part of the time.
3.2 Zilliz Cloud Dedicated (Enterprise) promises that the service availability is not less than 99.95%
Zilliz Cloud Dedicated (Enterprise) promises that the Downtime is ≤ 21.9 minutes in a month, calculated as 730 hours (30.41 days) per month. If the above service availability commitment is not met, you can obtain compensation in accordance with Article 4 of this Agreement.
4. Compensation Plan
4.1 Compensation Standard
Zilliz Cloud provides compensation based on monthly service availability. The credit of compensation is calculated as a percentage of the monthly service fee paid for the service that does not meet the service availability commitment as shown in the table below.
Compensation is limited to vouchers used to purchase Zilliz Cloud products, and the total credits shall not exceed the monthly service fee you paid for Zilliz Cloud products (excluding fees deducted by vouchers) when the service availability commitment was not met for promise.
| Service Availability | Service Credits |
|---|---|
| Less than 99.9% but equal to or greater than 99.5% | 10% |
| Less than 99.5% but equal to or greater than 95.0% | 25% |
| Less than 95.0% | 100% |
Given that you have purchased two clusters A and B in one month, the Service Availability is to be calculated for each of these clusters as follows:
- For cluster A which serves 10,000 minutes,
- The Service Availability is above 99.95% if the Service Downtime is less than 5 minutes.
- The Service Availability is less than 99.95% but equal to or greater than 99.5% if the Service Downtime is between 5 minutes and 50 minutes.
- For cluster B which serves a full month,
- The Service Availability is above 99.95% if the Service Downtime is less than 21.9 minutes.
4.2 Customer Obligations
To be eligible to claim service credits, you must submit a claim by opening a ticket on the Platform or send us an email at support@zilliz.com by the end of the month next to the one in which Service Downtime occurs. The claim must include
- The exact words "SLA Credit Request" in the subject line
- Information necessary for us to validate your claim, including but not limited to
- Dates and time of each Service Downtime you are claiming
- Names of the Cloud Provider and Deployed Region of your database affected by the Service Downtime
- Descriptions of your attempts to resolve the Service Downtime at the time of its occurrence
- Logs of the requests sent to your database within the Service Downtime with any sensitive information being removed or replaced by asterisks.
We will evaluate your claims and determine in good faith whether you shall be entitled to the Service Credits. We will issue any entitled Service Credits within the month next to the one in which your claim is received.
5. Exclusions
The service commitment does not apply to any unavailability, suspension, or termination of our services caused by
- Factors outside of our reasonable control, including any force majeure events or problems beyond the demarcation point of the Platform
- Any voluntary actions or inactions from you or any third party
- Insufficient compute and storage capacity provisioned to process your requests
- Credential leaks or improper authorization resulted from improper operations
- Failure to be in full compliance with the procedures and suggestions in the user guides and other product documents
- Services, hardware, or software provided by a third party, such as cloud platform services (e.g., AWS, GCP) on which Zilliz Cloud runs
- Zilliz Cloud’s Public Preview version and Private Preview version