[
https://issues.apache.org/jira/browse/GEODE-1187?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15232431#comment-15232431
]
ASF subversion and git services commented on GEODE-1187:
--------------------------------------------------------
Commit 5ba923e3ff2b682e51e4b69169292bde1065c46e in incubator-geode's branch
refs/heads/feature/GEODE-17-2 from [~bschuchardt]
[ https://git-wip-us.apache.org/repos/asf?p=incubator-geode.git;h=5ba923e ]
GEODE-1187 a server launched by Gfsh deletes PID file during auto-reconnect
ServerLauncher needed to be made aware of auto-reconnect. I changed
getCache() to look for a new, reconnected cache and I changed isWaiting()
to see if the current cache is in a reconnecting state.
I also changed the shutdown logic to cancel reconnect attempts if the
cache is in that state when a stop is requested.
> If a server launched by Gfsh goes into auto-reconnect the server's PID file
> is deleted
> --------------------------------------------------------------------------------------
>
> Key: GEODE-1187
> URL: https://issues.apache.org/jira/browse/GEODE-1187
> Project: Geode
> Issue Type: Bug
> Components: gfsh
> Reporter: Bruce Schuchardt
> Assignee: Bruce Schuchardt
> Fix For: 1.0.0-incubating.M2
>
>
> If a server launched by gfsh is kicked out of the distributed system and goes
> into auto-reconnect its process-ID file will disappear. This can cause
> monitoring programs to think that the process has failed and initiate
> creation of a new process.
> The problem seems to be that the gfsh ServerLauncher class has no checks for
> a reconnecting server. The launcher stops and deletes the PID file but the
> Reconnect thread and, subsequently, the Acceptor thread for the cache server
> keep the JVM from exiting.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
