-
Notifications
You must be signed in to change notification settings - Fork 14.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add rpc server to helm chart #38549
Add rpc server to helm chart #38549
Conversation
4c78c41
to
c49b449
Compare
Agree |
Maybe we rename rpcServer section “_rpcServer” until release? |
Why not. It will also be disabled by default so no big deal and we still have the feature flag |
Just to confirm.... given that it's behind feature flag anyway, would it be sufficent to just call it out as "private" in the docs? Feels like a stretch. @jedcunningham thought? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Let's prefix with _
, or maybe _aip44RpcServer
? I'm hesitant of a docs only change.
And yes the component wouldn't start, but it would go into crashloop. That said, it seems like we have a miss on our chart integration tests. A component being stuck in crashloop / never actually running should surely be a failure, no?!
a8302a1
to
bb2a928
Compare
…experimental / dev only
Previously you would not see the error message; you'd only see that the command failed. Now you'll see it.
…experimental / dev only
…experimental / dev only
Did it. Went with When we make it "public" then we'll have to change it both at top level and in .Values.ports._rpcServer. But that should be easy to catch since it will show up in a find / replace. I think this is should be ready to go now, presuming tests pass now, especially now that we made it "private". I added comments and newsfragment. |
148a60d
to
3eead4c
Compare
My knowledge in Kubernetes is too limited to review this PR :/ I'll leave it to better eyes than mine |
I think we should have some docs to explain about this capability (and mark it as experimental in the docs as well) |
Hi @eladkal Are you talking about the helm chart docs? Or the regular airflow docs? I looked in helm chart docs and don't see anything analogous for other services. E.g. there's no "triggerer" section or "scheduler" section. Is there a particular place you think an RPC server section would make sense? |
Yes I am talking about the helm chart docs. The reason I am raising this is you added newsfragmant that essentially is a new feature.. it's normally accompanied with docs |
I understand. Yeah, it's not released yet though. Maybe I should remove the newsfragment? Cus this is just private and for dev use only. So maybe it's the wrong time to add a newsfragment. The AIP is not near complete yet and actually, some of the comments on dev list recently indicate that now with 3.0 on the horizon, we might not actually ever complete AIP-44. But anyway, typically in helm chart, we don't document the service itself. That's left to the core docs. But the core docs have not yet been written (and reasonably so, I think, because we're not yet very near to completion and especially now with the uncertainty). So I think it's OK to leave undocumented for now. But I will remove the newsfragment because I think that's a reasonable thing to do. Let me know if you have additional concerns. |
I call the component "rpc server" beacause the there isn't really just one "internal API" -- the other one is the one used by the webserver.
I think we might want to consider renaming it at the CLI also.
Ideally, also we could somehow mark this (rpc server) private in the helm chart until AIP-44 is "done"