jrmccluskey commented on PR #17782: URL: https://github.com/apache/beam/pull/17782#issuecomment-1142492381
The runner *should* never call TrySplit(0.0) but it's theoretically possible, so proper handling in the case it does should be a good precaution. Returning a ProcessContinuation without trying to claim any work is definitely an anti-pattern as far as streaming SDFs are concerned. I'm confused by the distinction between TryClaim(i) and isDataAvailable(i) though, as in both cases once the position is claimed in the RTracker it won't be able to be claimed after the checkpoint. -- 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]
