dlmarion commented on PR #4787: URL: https://github.com/apache/accumulo/pull/4787#issuecomment-2269652821
> > I looked at how we could test for a duplicate FateId in the ConditionalMutationWriter. I didn't see an easy way to do that since the FateId is in the row. If we could do that, then we could set a status for duplicate FateId, and continue to retry, and then fail on REJECTED or UNKNOWN. > > As it works now, it is testing for a duplicate FateId (while not directly looking at the FateId). It will be REJECTED if there is already a TStatus set for the FateId, and a TStatus is always set on creation of the FateId and persists throughout the existence of the FateId. Maybe I'm misunderstanding you? > > Maybe there is no issue with both stores retrying infinitely, but what would be the problem with both stores retrying only a finite number of times? This could be a larger than 5, but still finite. The probability of collision would be, for all intents and purposes, impossible. If it fails after these attempts, then there is a bigger problem and not a problem with collisions I think I mis-understood your comment in relation to the code. You are saying that we should continue to retry on REJECTED, but throw an exception on UNKNOWN? -- 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]
