[
https://issues.apache.org/jira/browse/HDFS-3360?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13268141#comment-13268141
]
J.Andreina commented on HDFS-3360:
----------------------------------
Configured "dfs.namenode.checkpoint.dir" and
"dfs.namenode.checkpoint.edits.dir" to more than 6 comma seperated directories
with 5-6 levels in each directory SNN gets shutdown without throwing any
exception.
dfs.namenode.checkpoint.dir=
"/home/hadoop/hadoop-root/check/dfs/dir1,/home/hadoop/hadooproot/check/dfs/dir2,/home/hadoop/hadoop-root/check/dfs/dir3,/home/hadoop/hadoop-root/check/dfs/dir4,/home/hadoop/hadoop-root/check/dfs/dir5,/home/hadoop/hadoop-root/check/dfs/dir6,/home/hadoop/hadoop-root/check/dfs/dir7"
SNN gets shutdown without throwing any exception.But when configured to less
than or equal to 6 comma seperated values with 5-6 levels in each directory SNN
start up is fine .
The same behavior is observed with dfs.name.dir.But it throws the following
exceptions and NN start up fails.
NN logs
=======
{noformat}
java.lang.IllegalStateException: Variable substitution depth too large: 20
${hadoop.tmp.dir}/dfs/name1,${hadoop.tmp.dir}/dfs/name2,${hadoop.tmp.dir}/dfs/name3,${hadoop.tmp.dir}/dfs/name4,${hadoop.tmp.dir}/dfs/name5,${hadoop.tmp.dir}/dfs/name6,${hadoop.tmp.dir}/dfs/name7
{noformat}
> SNN gets Shutdown if the conf "dfs.namenode.checkpoint.dir" and
> "dfs.namenode.checkpoint.edits.dir" is configured to more than 6 comma
> seperated values with 3-4 level in each directories
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-3360
> URL: https://issues.apache.org/jira/browse/HDFS-3360
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: name-node
> Affects Versions: 2.0.0
> Reporter: J.Andreina
> Priority: Minor
> Fix For: 2.0.0, 3.0.0
>
>
> Configured "dfs.namenode.checkpoint.dir" and
> "dfs.namenode.checkpoint.edits.dir" to more than 6 comma seperated
> directories
> Started NN,DN,SNN
> Secondary Namenode gets shutdown without throwing any exception
> But the descriptions says that "If this is a comma-delimited list of
> directories then the image is
> replicated in all of the directories for redundancy."
> SNN logs
> ========
> {noformat}2012-04-26 13:08:37,534 INFO
> org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode: STARTUP_MSG:
> /************************************************************
> STARTUP_MSG: Starting SecondaryNameNode
> STARTUP_MSG: host = HOST-xx-xx-xx-xx/xx.xx.xx.xx
> STARTUP_MSG: args = []
> STARTUP_MSG: version = 2.0.0-SNAPSHOT
> STARTUP_MSG: build = -r ; compiled by 'isap' on Fri Apr 20 09:10:53 IST
> 2012
> ************************************************************/
> 2012-04-26 13:08:38,728 INFO org.apache.hadoop.metrics2.impl.MetricsConfig:
> loaded properties from hadoop-metrics2.properties
> 2012-04-26 13:08:38,861 INFO
> org.apache.hadoop.metrics2.impl.MetricsSystemImpl: Scheduled snapshot period
> at 10 second(s).
> 2012-04-26 13:08:38,861 INFO
> org.apache.hadoop.metrics2.impl.MetricsSystemImpl: SecondaryNameNode metrics
> system started
> 2012-04-26 13:08:39,176 INFO
> org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode: SHUTDOWN_MSG:
> /************************************************************
> SHUTDOWN_MSG: Shutting down SecondaryNameNode at HOST-xx-xx-xx-xx/xx.xx.xx.xx
> ************************************************************/{noformat}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
