yuqi1129 commented on PR #11892:
URL: https://github.com/apache/gravitino/pull/11892#issuecomment-4875609441
The recent frequency is from the Lance repair-on-load path being exercised
concurrently, especially by Spark loads in ITs. Two requests can load the same
declared or empty table and both try to persist the discovered schema/version
through optimistic-locked `EntityStore.update`. One update succeeds; the loser
sees `IOException("Failed to update the entity")`, which was previously
surfaced as HTTP 500 even though the table was already consistent. This PR
makes that lost CAS retry against the latest entity instead of failing the load.
--
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]
