Store: fix error handling on limits

[douglascamata]

• I added CHANGELOG entry for this change.
• Change is not relevant to the end user.

Changes

I noticed from running Thanos Store in production that limits were being hit quite often and this triggered an alert based on percentage of failed requests. A deep investigation allowed me to identify that the grpc codes for requests hitting limits made no sense: they were mostly "Unknown" and "Aborted". This motivated me to fix such logic.

• Removed a test implementation of the series and samples limiter factories used in a test to test its own behavior.
• Now such test uses the real implementation of such limiters.
• Fixed the error handling from the limiters, allowing a proper gRPC status code to be returned when limits are hit: ResourceExhausted.

Verification

E2E tests changed and passing.

[squat]

Nice find

[douglascamata]

As always, I forget to goimports my files 😭

Fixed in 60cf9a0.

[fpetkovski]

Thanks, do we need to fix this in other stores?

[douglascamata]

@fpetkovski from what I checked just now, yes. I can follow up with a PR fixing things around the limitedStoreServer. 😃