remove cpu limits follow up

[deer-wmde]

follow up to #2050

These are helmfile values only, so should be valid and mostly no-op (apart from staging superset because that value was not null). As far as I can see there is one limit left for the argo deployed redis but maybe we should touch that outside of this PR, as I'm currently not sure what brought us this ping-ponging back and forth between removing them and reverting again.

[tarrow]

So I did some digging from mattermost:

Tom
15:49
because then that re-introduced limits on workloads that weren't there before

15:50
it should definitely be re-applied btw

15:50
just after some chart updates to remove the hardcoded limits

I'm assuming it is not a no-op on charts that set default cpu limits; in this case the null was explicitly unsetting them

[deer-wmde]

ahh good catch, in that case we should check the charts first.

[tarrow]

I looked a bit today and got tempted to clean some up; I think we definitely have issues with queryservice for example (https://github.com/wbstack/charts/blob/main/charts/queryservice/values.yaml#L46)