[Optimization] boltdb readpath: reduce the code block affected by the lock tablesMtx in the TableManager.query() function

[liguozhong]

What this PR does / why we need it:
[Optimization] boltdb: reduce the code block affected by the lock tablesMtx in the TableManager.query() function.

util.DoParallelQueries(ctx, table, queries, callback) , free lock.
@sandeepsukhani boltdb issue.

Which issue(s) this PR fixes:
Fixes #

Special notes for your reviewer:

Checklist

• Documentation added
• Tests updated
• Add an entry in the CHANGELOG.md about the changes.

[sandeepsukhani]

Thanks for the PR! The changes look good to me but I doubt it would help much with the write latency since the read path is using a shared lock.

[sandeepsukhani]

@liguozhong this PR might help somewhat #5830

[liguozhong]

Thanks for the PR! The changes look good to me but I doubt it would help much with the write latency since the read path is using a shared lock.

Thanks for the PR! The changes look good to me but I doubt it would help much with the write latency since the read path is using a shared lock.

I just started reading boltdb code, trying to find lock related code, check if there are optimized code blocks.
This pr is because I saw it during the viewing process.
I'm still trying to find performance optimization points for writes.
We're working on switching from Cassandra to boltdb-shipper, which is a huge change for us, and we hope our team will be more familiar with the boltdb-shipper part of the code before switching to boltdb-shipper from cassandra.

[liguozhong]

@liguozhong this PR might help somewhat #5830

👍，The code of this PR（https://github.com/grafana/loki/pull/5830） would be helpful. After the PR is merged, we will release boltdb-shipper again to observe the optimization effect.