[
https://issues.apache.org/jira/browse/HDFS-13175?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16409053#comment-16409053
]
Anu Engineer commented on HDFS-13175:
-------------------------------------
[~yzhangal] Thanks for the comment and sharing the JIRA. That is precisely a
situation where this assertion can trigger. However, I do agree that
diskBalancer can be more graceful than fail via an assertion. If the capacity
is lesser than used space, the basic notions of space allocation are violated.
So it is appropriate that diskBalancer fails.
> Add more information for checking argument in DiskBalancerVolume
> ----------------------------------------------------------------
>
> Key: HDFS-13175
> URL: https://issues.apache.org/jira/browse/HDFS-13175
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: diskbalancer
> Affects Versions: 3.0.0
> Reporter: Lei (Eddy) Xu
> Assignee: Lei (Eddy) Xu
> Priority: Minor
> Fix For: 3.1.0, 3.0.2
>
> Attachments: HDFS-13175.00.patch, HDFS-13175.01.patch
>
>
> We have seen the following stack in production
> {code}
> Exception in thread "main" java.lang.IllegalArgumentException
> at
> com.google.common.base.Preconditions.checkArgument(Preconditions.java:72)
> at
> org.apache.hadoop.hdfs.server.diskbalancer.datamodel.DiskBalancerVolume.setUsed(DiskBalancerVolume.java:268)
> at
> org.apache.hadoop.hdfs.server.diskbalancer.connectors.DBNameNodeConnector.getVolumeInfoFromStorageReports(DBNameNodeConnector.java:141)
> at
> org.apache.hadoop.hdfs.server.diskbalancer.connectors.DBNameNodeConnector.getNodes(DBNameNodeConnector.java:90)
> at
> org.apache.hadoop.hdfs.server.diskbalancer.datamodel.DiskBalancerCluster.readClusterInfo(DiskBalancerCluster.java:132)
> at
> org.apache.hadoop.hdfs.server.diskbalancer.command.Command.readClusterInfo(Command.java:123)
> at
> org.apache.hadoop.hdfs.server.diskbalancer.command.PlanCommand.execute(PlanCommand.java:107)
> {code}
> raised from
> {code}
> public void setUsed(long dfsUsedSpace) {
> Preconditions.checkArgument(dfsUsedSpace < this.getCapacity());
> this.used = dfsUsedSpace;
> }
> {code}
> However, the datanode reports at the very moment were not captured. We should
> add more information into the stack trace to better diagnose the issue.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
