[
https://issues.apache.org/jira/browse/HDFS-13658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16516626#comment-16516626
]
Xiao Chen commented on HDFS-13658:
----------------------------------
Thanks for working on this Kitti! Looks good in general, some comments:
- Changing default fsck output is considered incompatible, see -HDFS-11259- /
HDFS-7933 for some examples. For this one, I think we can add a flag to fsck
and only calculate/output when the flag is provided
- {{LowRedundancyBlocks#incrementBlockStat}}, isStriped is already checked in
the if statement.
- Why is {{LowRedundancyBlocks#oneReplicaBlocks}} a set than a long? holding
all the blockinfo may be expensive.
- Need more tests to cover a few more scenarios: 1 replica, rf=1, should not
report; 1 replica, rf=3, should report; 2 replicas, rf=3, should not report.
Ideally we should do this in unit tests which would run faster than a
minidfscluster, see {{TestLowRedundancyBlockQueues}} for example.
- Good to add test coverage in {{TestNameNodeMetrics}} too
> fsck, dfsadmin -report, and NN WebUI should report number of blocks that have
> 1 replica
> ---------------------------------------------------------------------------------------
>
> Key: HDFS-13658
> URL: https://issues.apache.org/jira/browse/HDFS-13658
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs
> Affects Versions: 3.1.0
> Reporter: Kitti Nanasi
> Assignee: Kitti Nanasi
> Priority: Major
> Attachments: HDFS-13658.001.patch, HDFS-13658.002.patch,
> HDFS-13658.003.patch, HDFS-13658.004.patch
>
>
> fsck, dfsadmin -report, and NN WebUI should report number of blocks that have
> 1 replica. We have had many cases opened in which a customer has lost a disk
> or a DN losing files/blocks due to the fact that they had blocks with only 1
> replica. We need to make the customer better aware of this situation and that
> they should take action.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
