xBis7 commented on code in PR #5651:
URL: https://github.com/apache/ozone/pull/5651#discussion_r1441688656
##########
hadoop-ozone/recon/src/main/java/org/apache/hadoop/ozone/recon/fsck/ContainerHealthStatus.java:
##########
@@ -78,6 +100,14 @@ public int getReplicationFactor() {
}
public boolean isHealthy() {
+ return containerReplicaCount.isHealthy();
+ }
+
+ public boolean isSufficientlyReplicated() {
+ return containerReplicaCount.isSufficientlyReplicated();
+ }
+
+ public boolean isProperlyReplicated() {
Review Comment:
@sumitagrawl I couldn't figure out how to call this method. The original
name was `isHealthy` and I just wanted to change that, to avoid having a
different meaning here than the SCM has.
> We can rename method as isHealtyReplication() with above 2 condition,
That makes sense to me. I'll rename it.
> Can see some comment in code related to this point,
org.apache.hadoop.hdds.scm.SCMCommonPlacementPolicy#validateContainerPlacement
I added that comment in a previous PR, to explain that when having
`over-replication`, we shouldn't also have `mis-replication`. Although,
`over-replication` is an unhealthy state, `mis-replication` refers to not
having enough racks or having an uneven replica distribution per rack. Having
too many replicas in a rack, isn't an issue of `mis-replication` but just
`over-replication`.
> I think over-replication can be considered as healthy only (may not need
check delta=0).
With the above comment, it doesn't mean that when a container is
over-replicated then it's entirely healthy, we still need to fix
over-replication(disk space issue, etc.).
--
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]
