Iterator Perfomance Improvements #3667
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
joe-elliott
requested review from
annanay25,
mdisibio,
mapno,
kvrhdn,
zalegrala,
electron0zero,
ie-pham and
stoewer
as code owners
|
Gains due to the int64 -> int32 RowNumber conversion: Tag ValuesTraceQL SearchTraceQL Metrics |
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
joe-elliott
commented
May 13, 2024
| @@ -668,7 +668,7 @@ func mapSpanAttr(e entry) traceql.Static { | |||
| case columnPathSpanDuration: | |||
| return traceql.NewStaticDuration(time.Duration(e.Value.Int64())) | |||
| case columnPathSpanName: | |||
| return traceql.NewStaticString(e.Value.String()) | |||
| return traceql.NewStaticString(unsafeToString(e.Value.ByteArray())) | |||
There was a problem hiding this comment.
these unsafeToString calls massively reduce allocations b/c it prevents a clone on the string. The danger is in parquet-go reusing the backing buffer in a way that causes the value of the string to change.
Should we clone those strings that pass the test here:
i.e. only clone those strings that are passed up to the caller?
There was a problem hiding this comment.
You should be able to use the interner, so strings are cloned only once.
There was a problem hiding this comment.
oh! i forgot about the interner in the SyncIterator which actually does make a copy of the values returned here. so this unsafeToString call should be perfectly fine.
it's just mimicking this pattern already present in traceql search:
https://github.com/grafana/tempo/blob/main/tempodb/encoding/vparquet3/block_traceql.go#L2265
Signed-off-by: Joe Elliott <number101010@gmail.com>
Signed-off-by: Joe Elliott <number101010@gmail.com>
3 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
What this PR does:
A series of performance improvements that came from investigating the perf on tag value search.
GenericPredicateand included a note to not use it. There is one predicate still using this, but I would like to remove them all.GenericPredicatewas creating not inlineable function calls to do simple comparisons.unsafeToStringin the tag filters.make gen-parquet-queryto replace the usage ofGenericPredicate. This is included inmake vendor-checkso CI will confirm that the template and code do not drift.SyncIteratorInstrumentedPredicate.FloatBetweenPredicate. It was not usedTag Values Benches
TraceQL Search Benches
TraceQL Metrics Benches
Checklist
CHANGELOG.mdupdated - the order of entries should be[CHANGE],[FEATURE],[ENHANCEMENT],[BUGFIX]