[
https://issues.apache.org/jira/browse/HDFS-10999?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15963842#comment-15963842
]
Takanobu Asanuma commented on HDFS-10999:
-----------------------------------------
Thank you for this great work, [~manojg]! Although I haven't reviewed all the
detailed implementations yet, I'd like to add some comments.
* How about registering the new beans separately?
{code:title=FSNamesystem.java}
StandardMBean bean = new StandardMBean(this, FSNamesystemMBean.class);
StandardMBean replicaBean = new StandardMBean(this,
ReplicatedBlocksStatsMBean.class);
StandardMBean ecBean = new StandardMBean(this, ECBlockGroupsStatsMBean.class);
mbeanName = MBeans.register("NameNode", "FSNamesystemState", bean);
mbeanName = MBeans.register("NameNode", "ReplicatedBlocksState", replicaBean);
mbeanName = MBeans.register("NameNode", "ECBlockGroupsState", ecBean);
{code}
In this case, JMX is like below. This would be useful for administrators and
monitoring tools.
{noformat}
{
"name" : "Hadoop:service=NameNode,name=ECBlockGroupsState",
"modelerType" : "org.apache.hadoop.hdfs.server.namenode.FSNamesystem",
"ECLowRedundancyBlocksStat" : 0,
"ECCorruptBlocksStat" : 0,
"ECMissingBlocksStat" : 0,
"ECBlocksBytesInFutureStat" : 0,
"ECPendingDeletionBlocksStat" : 0
}
...
{
"name" : "Hadoop:service=NameNode,name=ReplicatedBlocksState",
"modelerType" : "org.apache.hadoop.hdfs.server.namenode.FSNamesystem",
"UnderReplicatedBlocksStat" : 0,
"CorruptReplicaBlocksStat" : 0,
"MissingReplicaBlocksStat" : 0,
"MissingReplicaOneBlocksStat" : 0,
"ReplicaBlocksBytesInFutureStat" : 0,
"PendingDeletionReplicaBlocksStat" : 0
}
{noformat}
* IMHO I prefer ec-related terminology is more look like its of fsck.
{noformat}
ECLowRedundancyBlocks -> LowRedundancyECBlockGroups
ECCorruptBlocks -> CorruptECBlockGroups
ECMissingBlocks -> MissingECBlockGroups
ECPendingDeletionBlocks -> PendingDeletionECBlockGroups
{noformat}
* From the earlier discussion, separating {{PendingReconstructionBlocks}} is
important. We can do this in a follow-on jira.
I will review this patch more deeply. Thanks again!
> Introduce separate stats for Replicated and Erasure Coded Blocks apart from
> the current Aggregated stats
> --------------------------------------------------------------------------------------------------------
>
> Key: HDFS-10999
> URL: https://issues.apache.org/jira/browse/HDFS-10999
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: erasure-coding
> Affects Versions: 3.0.0-alpha1
> Reporter: Wei-Chiu Chuang
> Assignee: Manoj Govindassamy
> Labels: hdfs-ec-3.0-nice-to-have, supportability
> Attachments: HDFS-10999.01.patch
>
>
> Per HDFS-9857, it seems in the Hadoop 3 world, people prefer the more generic
> term "low redundancy" to the old-fashioned "under replicated". But this term
> is still being used in messages in several places, such as web ui, dfsadmin
> and fsck. We should probably change them to avoid confusion.
> File this jira to discuss it.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
