Set default concurrency limit based on urgency & max threads
What does this MR do and why?
Set a default concurrency_limit
for all workers based on their urgency and maximum possible threads in the shard. The maximum threads is derived from max pods * concurrency
of the shard.
The max pods is retrieved from a new SIDEKIQ_MAX_REPLICAS
environment variable is added in gitlab-org/charts/gitlab!4348.
Proposed percentage based on urgency gitlab-com/gl-infra/data-access/durability/team#215 (comment 2571342290):
- High: 35%
- Low: 25%
- Throttled: 15%
References
gitlab-com/gl-infra/data-access/durability/team#215 (closed)
Also as 2nd prerequisite of gitlab-com/gl-infra/observability/team#3815 (closed)
Screenshots or screen recordings
Before | After |
---|---|
How to set up and validate locally
MR acceptance checklist
Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.
Edited by Marco Gregorius