dawidwys commented on pull request #18086:
URL: https://github.com/apache/flink/pull/18086#issuecomment-994656557


   > BTW, I also consider to let RocksDB state-backend to distinguish full 
snapshot with savepoint. Current full snapshot of RocksDB is actually a 
savepoint, which could be slow in production environment. This is another 
story, and I could create another ticket if you have simliar idea.
   
   Not sure, if I fully grasped the idea, but my intention is not to use the 
savepoint format for the "forced full snapshot". If we have incremental 
checkpoints enabled, the first checkpoint after `NO_CLAIM` would still be in 
the "incremental" format, it just won't reuse any sst files from the restored 
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]


Reply via email to