[
https://issues.apache.org/jira/browse/HDFS-16763?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17731825#comment-17731825
]
ASF GitHub Bot commented on HDFS-16763:
---------------------------------------
zhtttylz commented on code in PR #4868:
URL: https://github.com/apache/hadoop/pull/4868#discussion_r1227444307
##########
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/DFSConfigKeys.java:
##########
@@ -804,6 +804,9 @@ public class DFSConfigKeys extends CommonConfigurationKeys {
public static final long DFS_MOVER_MOVEDWINWIDTH_DEFAULT = 5400*1000L;
public static final String DFS_MOVER_MOVERTHREADS_KEY =
"dfs.mover.moverThreads";
public static final int DFS_MOVER_MOVERTHREADS_DEFAULT = 1000;
+ public static final String DFS_DATANODE_MOVER_MAX_NUM_CONCURRENT_MOVES_KEY =
+ "dfs.datanode.mover.max.concurrent.moves";
+ public static final int
DFS_DATANODE_MOVER_MAX_NUM_CONCURRENT_MOVES_DEFAULT = 10;
Review Comment:
Should the default value of "**dfs.datanode.mover.max.concurrent.moves**" be
consistent with "**dfs.datanode.balance.max.concurrent.moves**" at 100?
> MoverTool: Make valid for the number of mover threads per DN
> ------------------------------------------------------------
>
> Key: HDFS-16763
> URL: https://issues.apache.org/jira/browse/HDFS-16763
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: balancer & mover
> Reporter: Hongbing Wang
> Priority: Major
> Labels: pull-request-available
>
> When running the Mover tool, the number of mover threads per DN is always 1,
> resulting in very slow data movement.
> This JIRA fixes the problem which the current config is invalid.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
