harbor-helm
harbor-helm copied to clipboard
set proxy_request_buffering off
Fixes https://github.com/goharbor/harbor/issues/9217
I think this is quite an important change to make, thanks for submitting this @liuxu623 !
Another person also encountered the same issue and explained it: https://github.com/distribution/distribution/issues/2634
Essentially, we really do need to disable request buffering because pushing many large image layers at once can kill an ingress-nginx instance due to buffering too much data to disk. If ingress-nginx isn't run with ephemeral-storage limits (the default), then too much buffering can even destabilize a whole Kubernetes node due to disk pressure which triggers cascading evictions.
Seems the maintainers haven't noticed this PR; want me to recreate it for you @liuxu623 ?
This PR is being marked stale due to a period of inactivty. If this PR is still relevant, please comment or remove the stale label. Otherwise, this PR will close in 30 days.
This PR was closed because it has been stalled for 30 days with no activity. If this PR is still relevant, please re-open a new PR against main.