swamirishi commented on code in PR #7402:
URL: https://github.com/apache/ozone/pull/7402#discussion_r1848704439
##########
hadoop-hdds/container-service/src/main/java/org/apache/hadoop/ozone/container/common/impl/HddsDispatcher.java:
##########
@@ -276,7 +277,8 @@ private ContainerCommandResponseProto dispatchRequest(
getMissingContainerSet().remove(containerID);
}
}
- if (getMissingContainerSet().contains(containerID)) {
+ if (cmdType != Type.CreateContainer && !HddsUtils.isReadOnly(msg)
Review Comment:
We should allow writes to go through since we could be creating a RECOVERING
container done by replication manager. As of date that is a problem. Once a
container is marked missing we don't allow replication manager replicate the
same container on the datanode. Consider we have only 5 nodes in a cluster and
we lost one replica of an EC container because of volume failure. Now
reconstruction will fail forever because the create container command fired
would fail forever now.
--
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]
