reuvenlax commented on pull request #14718: URL: https://github.com/apache/beam/pull/14718#issuecomment-832094502
The general invariant for watermark-correct operators was that processing of an element with timestamp T does not produce a timestamp > T. This check was an attempt to extend that timer processing. Making this a hard check is maybe too strict, but I'm still not sure I understand the use case. Allowing future timestamps means that you have a watermark hold for timestamps T+delta, but you remove the watermark hold when the watermark reaches T. I guess we could allow it, but it still seems odd to me. On Tue, May 4, 2021 at 9:52 AM Kenn Knowles ***@***.***> wrote: > Nonetheless the check here doesn't have any basis in real requirements, > right? > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <https://github.com/apache/beam/pull/14718#issuecomment-832090563>, or > unsubscribe > <https://github.com/notifications/unsubscribe-auth/AFAYJVJYAUOUHL2ET4HD5FLTMAQ3PANCNFSM44CKPPKQ> > . > -- 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. For queries about this service, please contact Infrastructure at: [email protected]
