[
https://issues.apache.org/jira/browse/HADOOP-10638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14012654#comment-14012654
]
Manikandan Narayanaswamy commented on HADOOP-10638:
---------------------------------------------------
Testing done:
Installed NFS and updated the hadoop-daemon.sh from under /usr/lib/hadoop/sbin
with the patch included above.
Set up the following environment variables in /etc/defaults/hadoop-hdfs-nfs3
$HADOOP_PRIVILEGED_NFS_USER
$HADOOP_PRIVILEGED_NFS_LOG_DIR
$HADOOP_PRIVILEGED_NFS_PID_DIR
Then started and stopped NFS service successfully. The logs and pid file were
pushed to the expected directories with the expected file names.
> Updating hadoop-daemon.sh to work as expected when nfs is started as a
> privileged user.
> ----------------------------------------------------------------------------------------
>
> Key: HADOOP-10638
> URL: https://issues.apache.org/jira/browse/HADOOP-10638
> Project: Hadoop Common
> Issue Type: Bug
> Components: nfs
> Affects Versions: 2.4.0
> Reporter: Manikandan Narayanaswamy
> Assignee: Manikandan Narayanaswamy
> Labels: patch
> Attachments:
> 0001-Picking-the-right-pid-file-when-running-NFS-as-privi.patch
>
>
> When NFS is started as a privileged user, this change sets up required
> environment variables:
> HADOOP_PID_DIR = $HADOOP_PRIVILEGED_NFS_PID_DIR
> HADOOP_LOG_DIR = $HADOOP_PRIVILEGED_NFS_LOG_DIR
> HADOOP_IDENT_STRING = $HADOOP_PRIVILEGED_NFS_USER
> Also, along with the above, we also now collect ulimits for the right user.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
