xintongsong commented on pull request #15599:
URL: https://github.com/apache/flink/pull/15599#issuecomment-1019700632


   @galenwarren, sorry for the late response. I was OoO the past a few days.
   
   > The change I'd propose to help @sap1ens (or anyone else who used the 
pre-released code) would be to change 0 to 1 in the following lines in 
GSCommitRecoverableSerializer and GSResumeRecoverableSerializer
   
   We don't usually provide compatibility guarantees for pre-released codes. 
However, in this specific case, I see barely any cost in making lives of the 
pre-released code users easier. So +1 for changing the serializer version. I've 
already opened a hotfix PR #18409 for including `flink-gs-fs-hadoop` into 
`flink-dist`, and I can change the serializer version there.
   
   Concerning the authentication, do you have an idea how much effort the 2nd 
approach may require?
   - I think it would be ideal that we can fallback to `core-site.xml`, if it 
contains the credentials AND `GOOGLE_APPLICATION_CREDENTIALS` is not specified. 
This would be my top preference, if feasible.  
   - If the above approach requires too much effort / time (as we are 
approaching to the feature freeze date), I'd also be fine with documenting this 
as a known limitation (that `core-site.xml` does not work with the 
`RecoverableWriter` part), and address this as a follow-up issue in the next 
release cycle.
   - I would suggest not to exclude the `core-site.xml` approach from the 
documentation, unless we decide to deprecate / remove this approach for good. 
Otherwise, it might confuse users whether the approach is still available or 
not. And as far as I can see, there's not yet strong reasons for making such an 
incompatible change.


-- 
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: issues-unsubscr...@flink.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


Reply via email to