cnauroth commented on PR #26102:
URL: https://github.com/apache/flink/pull/26102#issuecomment-3216047004

   @rmetzger , sorry for not giving a status update sooner, but I had started a 
similar conversation in Hadoop on 7/23:
   
   https://lists.apache.org/thread/gy30bn6dm4qc3qwfdzdh2318m38yyvyn
   
   I then filed 
[INFRA-27071](https://issues.apache.org/jira/browse/INFRA-27071) to request a 
GCS bucket and credentials. The latest status in that bug is that infra has no 
presence on GCP and doesn't plan to use it, so it appears this is a non-starter 
at the moment.
   
   I agree that this PR is still useful in its current state, especially if 
people have the option to run manually before upgrades like FLINK-37328. Can we 
resume review and eventual merge?
   
   > Fully agree with you that this is a pretty risky approach, as it requires 
PRs to be carefully reviewed to not steal the credentials when merged.
   
   BTW as a side note, if we had gone ahead with this, then I would have 
advocated for setting it up as something committer-triggered (e.g. with a 
comment on the PR) only after careful review. Ideally the automation would also 
exchange service account credentials for a short-lived token with a tight 
downscoped access boundary, so even if something sneaks past us, at least it's 
not a powerful long-term credential.


-- 
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]

Reply via email to