surawut-jirasaktavee commented on issue #45877: URL: https://github.com/apache/superset/issues/45877#issuecomment-6529962844
Opened #45900. While testing I found two issues with the `identity_digest` approach (which I proposed above, and which #45895 also uses): 1. `identity_digest` is an HMAC keyed with `SECRET_KEY`. `docs/developer_docs/semantic-layer-cache-clear.md` says `SECRET_KEY` must not affect semantic cache identity, and `test_metadata_aliases_share_result_cache` fails with `CACHE_QUERY_BY_USER` on. During a rolling `SECRET_KEY` rotation, web and worker processes compute different keys, so async chart queries keep missing worker results. #45900 uses the server-authenticated `principal_id` instead, as datasets use the username. 2. Catching exceptions from the context lookup silently drops the user partition, so users would share results again. #45900 lets it raise, so it fails closed. The column-values cache has the same gap for datasets and semantic views; I opened #45898 for it. @mikebridge happy to go with whatever you prefer for the use case and approach. -- 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]
