You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Search for openjdk8-pipeline, see the double entry, and also note that the total failures at the end are "8", despite there being only one set of four failures in the double-printed jdk8 pipeline.
How it should work
It should count each failed build only once.
Next steps
To look at the code that fetches data from TRSS and figure out why we're receiving two sets of the same data, or why we're duplicating the data we receive.
The text was updated successfully, but these errors were encountered:
Likely related to the fact that ci.adoptopenjdk.net got renamed to ci.adoptium.net and we are currently monitoring both in TRSS, so double-counting. I have not yet deleted the configuration in TRSS that has us monitoring ci.adoptopenjdk.net, as I was testing whether I could just mark the monitoring to "No" but hitting a known bug adoptium/aqa-test-tools#775. I will remove the ci.adoptopenjdk.net monitoring completely which should fix this issue.
Summary
It appears that the system for reporting on the health of overnight builds is somehow counting failures twice.
Example: https://ci.adoptium.net/job/nightlyBuildAndTestStats_temurin/981//console
Search for openjdk8-pipeline, see the double entry, and also note that the total failures at the end are "8", despite there being only one set of four failures in the double-printed jdk8 pipeline.
How it should work
It should count each failed build only once.
Next steps
To look at the code that fetches data from TRSS and figure out why we're receiving two sets of the same data, or why we're duplicating the data we receive.
The text was updated successfully, but these errors were encountered: