supratimdeka commented on a change in pull request #1364: HDDS-1843.
Undetectable corruption after restart of a datanode.
URL: https://github.com/apache/hadoop/pull/1364#discussion_r320368137
##########
File path:
hadoop-hdds/container-service/src/main/java/org/apache/hadoop/ozone/container/common/impl/HddsDispatcher.java
##########
@@ -329,6 +336,21 @@ private ContainerCommandResponseProto dispatchRequest(
}
}
+ private void updateBCSID(Container container,
+ DispatcherContext dispatcherContext, ContainerProtos.Type cmdType) {
+ long bcsID = container.getBlockCommitSequenceId();
+ long containerId = container.getContainerData().getContainerID();
+ Map<Long, Long> container2BCSIDMap;
+ if (dispatcherContext != null && (cmdType == ContainerProtos.Type.PutBlock
Review comment:
an alternative implementation would be to ignore the cmdType.
For all requests, read the BCS ID from the map and compare it to the bcsid
from the container. if the map value is lower, then update it to the value from
the container.
Advantage:
updateBCSID function becomes de-coupled from the knowledge of which cmdType
changes the bcsid.
Disadvantage:
possibly more CPU cost, because every request will pay the cost of reading
the bcsid map. but it might be premature to assume that this additional cost
will be significant.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
