Service Level for Yandex Data Proc

This document defines the Service Level of Yandex Data Proc (“Service”) and is an integral part of Yandex.Cloud Customer Agreement (“Agreement”) and SLA available at: Capitalized terms used herein but not defined herein shall have the meanings set forth in Agreement and Linked Documents.

Service Level 99.50%

Compensation Amount:

Service Uptime Percentage in Accrual Period Compensation, % of Service Fee for Service in Accrual Period
< 99.50 % – 98.00 % 10.0 %
< 98.00 % – 90.00 % 25.0 %
less than 90.00 % 50.0 %

Definition of Service Unavailability:

The "Service Unavailability" is recorded when the Platform-side Error Rate exceeds 5 %.

The “Error Rate” is calculated as follows: the number of queries returning 500, 503 error codes or the "Internal Server Error" message is divided by the total number of valid queries for the same period of time. Queries that do not meet the Pending Query requirements are not taken into account in the formula.

The "Service Unavailability Period" means 10 consecutive minutes during which the Service Unavailability was recorded.

The “Valid Query” means a query that conforms with the Service documentation and returns a result indicating no error if the Service functions properly.

The "Pending Query" means the need for the Client to allow a time interval between queries to the server when an error occurs. According to the Pending Query rules, the timeout periods must be at least 1 second for the first repeated query and increase exponentially up to 32 seconds for each subsequent query.

The Service Level Agreement does not apply to cases where the Service Unavailability resulted from circumstances beyond the control of Yandex, including failures in external networks or the Client's failure to comply with the Instructions.


Yandex Services AG

Web address:

Date of publication: October 01, 2019

Effective date: October 01, 2019