[
https://issues.apache.org/jira/browse/HDFS-11947?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16042460#comment-16042460
]
Weiwei Yang commented on HDFS-11947:
------------------------------------
I could easily reproduce this
{code}
Configuration conf = new HdfsConfiguration();
conf.set(DFSConfigKeys.DFS_NAMENODE_LIFELINE_RPC_ADDRESS_KEY, "localhost:0");
try (MiniDFSCluster cluster = new
MiniDFSCluster.Builder(conf).numDataNodes(3).build()) {
cluster.waitActive();
}
{code}
Lifeline thread is started before
{noformat}
017-06-08 17:29:05,902 [main] WARN datanode.DataNode
(BPOfferService.java:getBlockPoolId(193)) - Block pool ID needed, but service
not yet registered with NN, trace:
java.lang.Exception at
org.apache.hadoop.hdfs.server.datanode.BPOfferService.getBlockPoolId(BPOfferService.java:193)
{noformat}
This is because lifeline thread was started before the DN-NN handshake, while
NamespaceInfo is not yet available. Uploaded a simple patch to fix this. Please
kindly review.
> When constructing a thread name, BPOfferService may print a bogus warning
> message
> ----------------------------------------------------------------------------------
>
> Key: HDFS-11947
> URL: https://issues.apache.org/jira/browse/HDFS-11947
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: datanode
> Reporter: Tsz Wo Nicholas Sze
> Assignee: Weiwei Yang
> Priority: Minor
>
> HDFS-11558 tries to get Block pool ID for constructing thread names. When
> the service is not yet registered with NN, it prints the bogus warning "Block
> pool ID needed, but service not yet registered with NN" with stack trace.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
