Service Level Agreement (“SLA”) Kubernetes Kapsule and Kosmos Dedicated Control Plane Service
This Service Level Agreement defines the different service level objectives that Scaleway undertakes to comply with in the context of the Covered Services, as well as their scope of application.
Definitions
“Covered Service”: means a “Kubernetes Dedicated Control Plane” provided by Scaleway, either in Kapsule or Kosmos type.
“Unavailability Period”: means the loss of external connectivity to the running cluster due to the loss of access to all of the Kubernetes API servers, without the possibility to launch a replacement cluster in any region for a duration of at least five (5) consecutive minutes.
Being specified that the following events shall in no circumstances be considered as an Unavailability Period:
- Any unavailability that occurs during a Maintenance Window; or
- Any failure of Kubernetes nodes or Kubernetes pods running on such nodes; or
- Any intermittent unavailability lasting less than five (5) consecutive minutes.
“Maintenance Window'': means a period of time during which Scaleway performs maintenance activities on the infrastructure (physical, network or software) upon which a Covered Service is based. This includes any updates of the Kubernetes API that may involve the disabling of a Covered Service.
“Monthly Availability Rate” or “MAR”: means the availability rate of a Covered Service for a given calendar month, determined by the following formula:
MAR = ((Tt-Tpit) / Tt) *100It being specified that:
- Tt: Total number of minutes in the relevant month (in case the Client partially runs a Covered Service during the course of a month, it is assumed that such Covered Service is 100% available for the portion of the relevant month that it was not running);
- Tpit: Number of minutes of unavailability recorded during every Unavailability Period that occurred in the relevant month.
Service Level Objectives “SLO”
Scaleway undertakes to comply with the following Service Level Objectives:
| Configuration | Monthly Availability Rate |
|---|---|
| Dedicated Control Plane (4, 8 or 16GB RAM) | 99.5% |
| Other configurations | No SLO |
The SLO described hereinabove shall in no circumstances apply to any components under Client's control, such as, but not limited to, any software and/or applications installed and used by the Client in connection with the Covered Services.
Compensation
In the event Scaleway does not comply with the SLO described hereinabove during a given calendar month, the Client is entitled to request a compensation, it being specified that (i) the Client shall have paid all invoices issued by and due to Scaleway in connection with the Services, (ii) the Client shall have notified Scaleway the non-respect of the SLO within the specified time limit; and (iii) the Client shall cooperate in good faith with Scaleway to resolve the issue, and in particular:
- by remaining available during the entire period necessary for resolving the issue encountered;
- by providing all the information reasonably available to the Client regarding the issue encountered;
- by carrying out any verification that may reasonably be necessary to resolve the issue encountered.
It is expressly agreed that the compensation granted by Scaleway constitutes an all-inclusive payment for all damages resulting from Scaleway’s failure to comply with the SLO and that the Client expressly waives its right to any other claim, complaint or action.
The compensation shall take the form of a voucher which shall be applied by Scaleway on the invoice issued for the Covered Service, following the month during which the Client’s request was issued.
The amount of the compensation granted by Scaleway is determined according to the following table and corresponds to a percentage of the monthly amount invoiced to the Client for the Covered Service that did not comply with the SLO during a given month. The Client acknowledges that this compensation cannot under any circumstances be reimbursed in cash.
| Monthly Availability Rate (“MAR”) | Compensation |
|---|---|
| Dedicated Control Plane (4, 8 or 16GB RAM) | Percentage of the monthly amount invoiced to the Client for the Covered Service that did not comply with the SLO |
| 99.0% < MAR < 99.50% | 10% |
| 95.0% < MAR < 99.0% | 25% |
| MAR < 95.0% | 50% |
Request for compensation
Client’s request for compensation must be issued no later than thirty (30) days following the month during which the applicable Covered Service did not comply with the SLO. Failing that, no compensation can be claimed by the Client with respect to such breach of the SLO.
In addition to its request, using any observability solution, the Client must provide Scaleway with log files (e.g. controller-manager) indicating the Unavailability Periods of the impacted Covered Service as well as the date and time at which they occurred. Client’s failure to comply with these requirements may result in the withdrawal of its right to compensation. In the event of a dispute regarding this SLA, Scaleway will take a decision in good faith based on the information provided by the Client as well as its own system logs, monitoring reports, configuration records and any other available information.
Maximum compensation
The amount of the compensation granted by Scaleway over the course of a given calendar month shall under no circumstances exceed fifty percent (50%) of the monthly amount invoiced to the Client for all the Covered Services that failed to comply with the SLO during the course of such month.
Exclusions
No compensation shall be granted to the Client if the failure to comply with the SLO is due to any of the following reasons:
- Force Majeure events;
- Act of a third party;
- Client’s fault or wrongful acts;
- Failure of the Client to fulfill its contractual obligations;
- Unavailability of a Covered Service occurring during a Maintenance Window;
- Misuse (or non-conforming use) of the Covered Service by the Client (such as but not limited to: updating cluster configuration, updating cluster version, disabling security groups, misconfiguration of VPC, RAM consumption above threshold, etc.);
- Usage of open source orchestration software that leads to repeated Kapsule or Kosmos cluster crashes or an inoperable Dedicated Control Plane (e.g., un-documented bugs in Kubernetes source code);
- Suspension of a Covered Service pursuant to the General Terms of Services or the applicable Specific Conditions;
- Internet connection issues;
- Failure of the Client to cooperate in good faith with Scaleway to restore the availability of a Covered Service;
- Services or functionalities in Beta test phase or which have not been made available in a “General Availability” phase;
- Services or functionalities for which there are no SLO explicitly described in the SLA.
Last modified: 18/09/2023