-
Notifications
You must be signed in to change notification settings - Fork 1.9k
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
WIP Code Coverage #2180
WIP Code Coverage #2180
Conversation
Thanks for working on this @codemzs . My understanding from @sharwell is that we can get the coverage test build from taking 10m during the test phase down to maybe 15m. But this depends upon future releases of Coverlet. (@sharwell correct me if I say anything wrong.) If we can get it down that far, might be worth taking Sam's advice elsewhere and putting it in main build. I don't have full context on what is going on behind the scenes. |
There are two main ways to get the speed improved:
@codemzs You could start by excluding the above three classes using |
@sharwell After excluding the classes you mentioned the build time with CC for Windows x64 debug went down to ~32 minutes from ~47 minutes. The build time without CC for the same platform is ~20 minutes. Link to build logs: https://dev.azure.com/dnceng/public/_build/results?buildId=75554 |
@codemzs Those numbers should be good enough to hold us over until coverlet is updated 👍 |
@sharwell Just to confirm, are you saying we should update the windows x64 debug leg of main CI build with the current coverlet nuget? If yes, I hope @TomFinley is cool too! :) |
@codemzs It could go in either leg. I'm saying this performance is enough that we can stop looking for |
@sharwell Either leg as in? We are sticking with just one CI right? |
Meaning it can go in MachineLearning-CodeCoverage or in MachineLearning-CI. If it goes in MachineLearning-CI, we just remove MachineLearning-CodeCoverage. |
No description provided.