-
Notifications
You must be signed in to change notification settings - Fork 433
perf(spanner): Use arenas to speed up queries that fetch many rows #15441
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
@scotthart could you take a look / do the gcbrun incantations |
|
/gcbrun |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #15441 +/- ##
=======================================
Coverage 93.04% 93.04%
=======================================
Files 2403 2403
Lines 219553 219644 +91
=======================================
+ Hits 204273 204368 +95
+ Misses 15280 15276 -4 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
I updated the code to fix the clang tidy warning |
|
/gcbrun |
|
We should merge |
|
Done, thanks |
scotthart
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@scotthart reviewed 11 of 12 files at r1, 1 of 1 files at r2, all commit messages.
Reviewable status:complete! all files reviewed, all discussions resolved (waiting on @andrewro-google)
|
/gcbrun |
|
/gcbrun |
Tested by running
multiple_rows_cpu_benchmark --project=${GOOGLE_CLOUD_PROJECT} --instance=${GOOGLE_CLOUD_CPP_SPANNER_TEST_INSTANCE_ID} --table-size=1000000 --maximum-channels=8 --maximum-threads=16 --iteration-duration=5 --samples=60 --experiment=select-string --use-only-clients=true --query-size=100000(larger --query-size than in the README). The data is noisy but seems like a ~20% win on CpuTime. I don't expect this to help much with queries that return few results. For the use case I care about (fetching a whole table), we were limited by CPU when using all the threads on a machine.We still have a copy when generating Row in PartialResultSetSource::NextRow() but we'd need a new API to avoid that.
This change is