You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently we do not set GOMAXPROCS and GOMEMLIMIT which means that the defaults will be used based on the host machine and not the container resources. This can have repercussions like CPU throttling due to Spegel using more CPU than was actually allocated.
Proposed solution to the problem
Good example for how to set this in the Helm chart based on resource limits.
Describe the problem to be solved
Currently we do not set GOMAXPROCS and GOMEMLIMIT which means that the defaults will be used based on the host machine and not the container resources. This can have repercussions like CPU throttling due to Spegel using more CPU than was actually allocated.
Proposed solution to the problem
Good example for how to set this in the Helm chart based on resource limits.
https://gitlab.com/gitlab-org/charts/gitlab-agent/-/merge_requests/104
Another blog post with an example.
https://metoro.io/blog/go-production-performance-gotcha-gomaxprocs
We should use the benchmarks to measure the performance to see if there is any difference when configuring these values.
The text was updated successfully, but these errors were encountered: