avoids unneeded meta index read #6024
Merged
+29
−22
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
CachableBlockFile.Reader.getBCFile() always read the RFile root block from the index cache even if would not use it. Modified the method to only read from the index cache when the data is actually needed. Noticed this while working on #6010, saw for every data block read from a rfile it would also read from the index cache.
Adjusted ScanTracingIT because before this change there was an rfile index cache read per rfile data block read. After this change the number of rfile index reads align w/ the number of rfiles opened, which in this case aligns with the number of tablets in the test tables.
The existing behavior was likely not a performance problem if the data was in cache, but could potentially cause thread contention and extra CPU usage. So its probably good to fix, also nice to make the tracing stats align better w/ expectations.