[
https://issues.apache.org/jira/browse/FALCON-442?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14016879#comment-14016879
]
Jean-Baptiste Onofré commented on FALCON-442:
---------------------------------------------
It's weird, in the bin/service-start.sh, we check if the PID doesn't already
exist and if the process is not there:
{code}
if [ -f $FALCON_PID_FILE ]; then
if kill -0 `cat $FALCON_PID_FILE` > /dev/null 2>&1; then
echo $APP_TYPE running as process `cat $FALCON_PID_FILE`. Stop it first.
exit 1
fi
fi
{/code}
I try to reproduce your issue and fix that.
> overwriting of PID while multiple falcon start
> -----------------------------------------------
>
> Key: FALCON-442
> URL: https://issues.apache.org/jira/browse/FALCON-442
> Project: Falcon
> Issue Type: Bug
> Components: general
> Affects Versions: 0.4, 0.5
> Environment: production
> Reporter: Samarth Gupta
>
> if falcon-start command is run on a already running falcon server, it goes
> and updates the PID that was stored during the previous falcon-star, however
> falcon keep running with the original PID.
> Due to this , when falcon-stop is run we get a missing process error, similar
> to below.
> {code}
> /usr/lib/falcon/falconServer/bin/service-stop.sh: line 36: kill: (19518) - No
> such process
> {code}
--
This message was sent by Atlassian JIRA
(v6.2#6252)
