This repository has been archived by the owner on Feb 20, 2023. It is now read-only.
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Outdated suggestions cannot be applied.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.
There is currently a bug in the oltpbench scripts where the latency numbers are converted into ints before being sent to the performance storage service. This causes problems with the Grafana visualizations because a number of the latency metrics end up being <1. This causes /0 errors as you can see here https://stats.noise.page/grafana/d/latency_trends/latency-trends?orgId=1
This PR fixes this issue and refactors the code by implementing reuse to give a single place to control the parsing of latency metric labels. This will be especially helpful if oltpbenchmark/oltpbench#338 gets merged (though it also works without that other PR).
I tested this code on both pipelines:
Terrier: http://jenkins.db.cs.cmu.edu:8080/blue/organizations/jenkins/testing-team%2Fterrier/detail/latency-int-bug/3/pipeline/375
Nightly: http://jenkins.db.cs.cmu.edu:8080/blue/organizations/jenkins/testing-team%2Fterrier-nightly/detail/latency-int-bug/5/pipeline/