[ https://issues.apache.org/jira/browse/HDFS-6101?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15020188#comment-15020188 ]
Walter Su commented on HDFS-6101: --------------------------------- bq. When there are 10 writers begin to writer at the same time, the policy will not allow some writers set up pipelines with 3 data nodes, due to the load factor of data nodes. It happens because in test we only start few DNs and write a lot files. In production It won't be a problem. I saw nodes be excluded by placement policy many times when I write tests for erasue-coded files, which writes to 9 DNs concurrently only for one file. So could you try {{conf.setBoolean(DFSConfigKeys.DFS_NAMENODE_REPLICATION_CONSIDERLOAD_KEY, false);}} and don't reduce the writers? > TestReplaceDatanodeOnFailure fails occasionally > ----------------------------------------------- > > Key: HDFS-6101 > URL: https://issues.apache.org/jira/browse/HDFS-6101 > Project: Hadoop HDFS > Issue Type: Bug > Reporter: Arpit Agarwal > Assignee: Wei-Chiu Chuang > Attachments: HDFS-6101.001.patch, HDFS-6101.002.patch, > HDFS-6101.003.patch, HDFS-6101.004.patch, HDFS-6101.005.patch, > TestReplaceDatanodeOnFailure.log > > > Exception details in a comment below. > The failure repros on both OS X and Linux if I run the test ~10 times in a > loop. -- This message was sent by Atlassian JIRA (v6.3.4#6332)