Hey @guwirth
We have implemented several tickets that will be available in the next release to bring the total analysis time (scanner + Compute Engine) closer to the prior performance.
SONAR-22998 - Optimize loading of active rules from the scanner
SONAR-24006 - Big regression in the Compute Engine step “Load quality profiles”
SONAR-23974 - Use enums for impacts in the scanner report to its reduce size
SONAR-24057 - Optimize querying of measures to only load required measures in memory
I don’t expect performance will return to the exact same – SonarQube has evolved in the last 2 years, it stores more data per issue, indexes more files… but hopefully these address the most severe performance degradations. I’ll be interested to see how your tests perform once the binaries are available!