Fix carbonserver render error race

[emadolsky]

It is a simple fix to #436.

Whenever there are multiple render requests approaching for the
same records, and the proceeding request gets an error, waiting
requests should get an error too.

It is a bit different from #431 because listener.prepareDataProto handles the 404 case and returns an empty response which will be handled correctly when it proceeds. So handling the error case would probably be the solution.

[deniszh]

Could you please check the fix, @bom-d-van ?

[bom-d-van]

yep, overall it looks good to me.

There is a DeepSource failure, @emadolsky could you check if you can fix it?
(not sure what's going on in golangci-lint.)

[deniszh]

Not sure about DeepSource - it complains about unused metric parameter - https://deepsource.io/gh/go-graphite/go-carbon/run/617e9c52-2797-47f7-90f9-e9b47844eb47/go/RVV-B0012 but lint error is clear, though - https://go-critic.com/overview.html#ifElseChain-ref
IMO replace if chain with switch case is good suggestion overall.

[emadolsky]

The metric name variable was not used in the function. The only use could be logging it. Also we could just get rid of it. But I was not sure which to choose. What do you think?

[bom-d-van]

The metric name variable was not used in the function. The only use could be logging it. Also we could just get rid of it. But I was not sure which to choose. What do you think?

logging seems good for me. approving it and re-running the jobs.

[bom-d-van]

The CodeQL ones look like the reaction of golang linter to the log4j incident. I'm reviewing them, probably gonna mark them as won't fix or false positive. please let me know if you disagree.