sbp opened a new issue, #637: URL: https://github.com/apache/tooling-trusted-releases/issues/637
When we have a check result cache hit, we still duplicate the previous result in the database even though we don't run the task. We do this because we look up check results by the current revision number, not by the hash of the file, partly because the hashes of filenames were not available until recently, and even now are only available by parsing attestation JSON data. Duplicating check results for every revision is not scaling very well even on my small local test instance: ``` $ sqlite3 state/database/atr.db 'SELECT COUNT(*) FROM checkresult;' 826628 ``` Therefore we should deduplicate. This is strongly related to #636, using caching on more checks and as early as possible. When there is a cache hit, we should just return. This would make cache hits very fast, because they would only require one database query and no writes. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
