ranger-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "suja s (JIRA)" <j...@apache.org>
Subject [jira] [Created] (RANGER-2229) Perform graceful terminate with retries before doing forceful kill for usersync and tagsync
Date Wed, 19 Sep 2018 06:05:00 GMT
suja s created RANGER-2229:
------------------------------

             Summary: Perform graceful terminate with retries before doing forceful kill for
usersync and tagsync
                 Key: RANGER-2229
                 URL: https://issues.apache.org/jira/browse/RANGER-2229
             Project: Ranger
          Issue Type: Bug
          Components: tagsync, usersync
    Affects Versions: 1.1.0
            Reporter: suja s


In ranger usersync and tagsync scripts, stop of the process is done forcefully using kill
-9.
Suggestion is to do a graceful shutdown with retries and do a force kill only if graceful
shutdown is not successful.

Stop section from ranger-tagsync-services.sh and ranger-tusersync-services.sh
{noformat}
lif [ "${action}" == "STOP" ]; then
 WAIT_TIME_FOR_SHUTDOWN=2
 NR_ITER_FOR_SHUTDOWN_CHECK=15
 if [ -f "$pidf" ] ; then
 pid=`cat $pidf` > /dev/null 2>&1
 echo "Found Apache Ranger TagSync Service with pid $pid, Stopping..."
 kill -9 $pid > /dev/null 2>&1
 sleep 1 #Give kill -9 sometime to "kill"
 if ps -p $pid > /dev/null; then
 echo "Wow, even kill -9 failed, giving up! Sorry.."
 else
 rm -f $pidf
 echo "Apache Ranger Tagsync Service pid = ${pid} has been stopped."
 fi
 else
 echo "Ranger Tagsync Service not running"
 fi
 exit;
{noformat}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message