smengcl commented on code in PR #5301:
URL: https://github.com/apache/ozone/pull/5301#discussion_r1338273163
##########
hadoop-ozone/interface-client/src/main/proto/OmClientProtocol.proto:
##########
@@ -1815,6 +1818,16 @@ message SnapshotPurgeRequest {
repeated string updatedSnapshotDBKey = 2;
}
+message SnapshotUpdateSizeRequest {
+ repeated SnapshotSize snapshotSize = 1;
+}
+
+message SnapshotSize {
Review Comment:
@aswinshakil Yup that can be a problem. But I'm worried that refactoring
would be much harder later on (when we implement this and want to merge update
SnapshotInfo operations).
Can we use a snapshotTableLock to synchronize the operations (e.g. snapshot
delete and snapshot update), so that the case you listed above won't happen?
The current approach relies on the current state of single-threaded
execution of Ratis Txs. And this is can also be a hidden problem when we enable
parallel Ratis Txs at some point. So a `SNAPSHOT_LOCK` for example inside of
`OMSnapshotUpdateSizeRequest#validateAndUpdateCache` would still be required.
Just like the `VOLUME_LOCK/BUCKET_LOCK` that we use in volume/bucket/key
operations.
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
