Grouper Users - Open Discussion List

[Chris Hyzer <>]

As long as it is a different JVM process, then you are all good.

Yes, you can share the same tomcat installation.  Yes, same value for 
TOMCAT_HOME and CATALINA_HOME, but different for CATALINA_BASE.  Have 
different conf, temp, work, webapps, logs dir.  In the conf, specify 
different ports.

Just to be clear, this is a java/tomcat productionizing issue.  If you share 
a tomcat, apps can affect each other, e.g. if one runs out of heap, I think 
the whole tomcat will have problems.  Restarting or redeploying should not 
affect other apps...  we have one apache that talks to all the tomcat so the 
URLs can be similar...

Thanks
Chris


-----Original Message-----
From: 

 
[mailto:]
 On Behalf Of Peter DiCamillo
Sent: Wednesday, April 17, 2013 10:31 AM
To: Chris Hyzer
Cc: 

Subject: Re: [grouper-users] Tomcat warning for c3p0

They are in the same tomcat, and I can try that. There's no need for hot 
redeploys. Do you mean completely separate installations of tomcat, or 
one tomcat but using different values for CATALINA_BASE ?

Peter

On 4/17/13 1:52 AM, Chris Hyzer wrote:
> Are you running the UI and WS in the same tomcat?  Can you try running in 
> their own tomcat?  I don't know why you would need to do that, but that is 
> how we run it at Penn :)  This isn't a dev env where you are doing hot 
> redeploys, right?
> 
> Thanks
> Chris
> 
> -----Original Message-----
> From: 
> 
>  
> [mailto:]
>  On Behalf Of Peter DiCamillo
> Sent: Wednesday, April 17, 2013 12:26 AM
> To: 
> 
> Subject: [grouper-users] Tomcat warning for c3p0
> 
> I'm trying to track down a problem which results in the Grouper UI 
> eventually becoming unresponsive. When I restart tomcat, it starts up 
> without warnings or errors. However, the first time a web services call 
> is made, this warning appears in catalina.out:
> 
> 2013-04-16 23:29:09,775: [ajp-8009-1] WARN 
> ActiveManagementCoordinator.attemptManageC3P0Registry(56) - A 
> C3P0Registry mbean is already registered. This probably means that an 
> application using c3p0 was undeployed, but not all PooledDataSources 
> were closed prior to undeployment. This may lead to resource leaks over 
> time. Please take care to close all PooledDataSources.
> 
> Is that something to be concerned about, and possibly the cause of my 
> problem? I haven't tried enabling debug logging yet, but the only other 
> unusual thing I see in the logs are some messages like this from time to 
> time:
> 
> 2013-04-16 23:29:20,927: [ajp-8009-1] WARN 
> WsStemLookup.retrieveStemIfNeeded(208) - Invalid query: WsStemLookup[
>    stemFindResult=INVALID_QUERY]
> 
> This is with Grouper 2.1.4.
> 
> Peter
>