enable configurable Kubernetes API rate limiting via env vars #1706
Conversation
There was a problem hiding this comment.
I think the goal of this change makes sense, but we'd need to plumb the env vars through the Provider schema for consistency. This would allow these values to be set on a per-Provider basis, and also documents the options in the API docs.
While we're making this change, are there any other k8s client settings that would make sense to expose?
|
@lblackstone I'm happy to look at trying to thread the parameters through the provider settings. Go isn't one of my primary languages right now, but I can give it a shot if you think that would be a good MVP. I felt like the env var first approach might be good to validate it before shipping it as part of the supported API. what do you think? |
|
@lblackstone hey! would you like me to plumb these through the provider schema or OK to use the env vars as an MVP for now? If the former, could you point out a couple key files that I should pay attention to when doing that plumbing, or an env var that currently affects the Kube client-go construction? |
|
@AaronFriel Hey, sorry for the delayed response here. (Have been busy working on the AWS Native provider launch!) I'm about to make some changes to the provider opts anyway, so I can pull in your changes while I'm at it. I'll tag you on the PR when it's ready. |
|
Added this feature in #1748, so I'll close this out now. Thanks for the idea and PR! |
|
Thank you! |
Proposed changes
Adds environment variables
PULUMI_KUBE_CLIENT_QPSandPULUMI_KUBE_CLIENT_BURSTto enable tuning Pulumi for large Kubernetes stacks with kube-apiservers that can handle the load.Related issues (optional)
Closes #1704