Re: [Wrapper-user] Wrapper 2.2.9.1 causing System.exit(1)

[Leif M. <le...@ta...>]

Nazim Nizar wrote:

> Hi,
>  
> I am running RMI server using wrapper 2.2.9.1 as service on Windows 
> 2000 Server. It uses log4j for logging. When I set root priority level 
> to debug in log4j config file (not wrapper config file), it terminates 
> abnormally while starting up. Works fine if I set the priority level 
> to info. I get the following in wrapper log after wrapper gives up:
>  
> INFO   | jvm 8    | 2003/03/02 15:44:57 | calling System.exit(1)

This line was output by the Wrapper as the last step in the Wrapper shutting
down the JVM.  I would need to know what is happening before this to give
you any clues as to what the problem could be.

> INFO   | jvm 8    | 2003/03/02 15:44:57 | 28470 [Thread-2] DEBUG 
> com.workforcesoftware.Policy.AllPolicies  - Loading Formula: _796BA2E4
> DEBUG  | wrapperp | 2003/03/02 15:44:57 | read a packet 107 : 0
> DEBUG  | wrapper  | 2003/03/02 15:44:57 | JVM signalled that it was 
> stopped.
> DEBUG  | wrapperp | 2003/03/02 15:44:57 | socket read no code (closed?).
> INFO   | wrapper  | 2003/03/02 15:44:57 | Wrapper Process has not 
> received any CPU time for 29 seconds.  Extending timeouts.

The Wrapper will kick out this warning message when its main loop has not
received any CPU for long periods of time.  In this case 29 seconds.
You are running a 4CPU machine?  I'm afraid I don't have access to one of
those, but it works fine with my 2CPU box.  What is the CPU usage while this
is happening?  Are all 4 CPUs pegged at 100% CPU?

> DEBUG  | wrapper  | 2003/03/02 15:44:57 | JVM Ping Failed.
> ERROR  | wrapper  | 2003/03/02 15:44:57 | JVM exited unexpectedly.
> DEBUG  | wrapper  | 2003/03/02 15:44:57 | JVM was only running for 97 
> seconds leading to a failed restart count of 5.
> FATAL  | wrapper  | 2003/03/02 15:44:57 | There were 5 failed launches 
> in a row, each lasting less than 300 seconds.  Giving up.
> FATAL  | wrapper  | 2003/03/02 15:44:57 |   There may be a 
> configuration problem: please check the logs.
> STATUS | wrapper  | 2003/03/02 15:44:58 | <-- Wrapper Stopped

At this point, I am not sure what the problem is, but my guess is that 
your application
is kicking out so much log output that it is eating all of the CPU.   If 
either the Wrapper
or its JVM does receive enough CPU to communicate with each other for an 
extended
period of time, then the Wrapper will assume a problem and restart the 
JVM.  It should
be in the logs above what you sent me, but my guess is that the Java 
side of the
Wrapper gave up waiting for the Wrapper to ping it and shut itself down.

You might want to think about redirecting all of that log output to a 
file.  That should clear
up the problem as it is much less CPU intensive than sending output to 
the console.

Another option is to extend the wrapper.ping.timeout or 
wrapper.startup.timeouts
depending on where your application is exiting.

> Is there something related to above fixed in 3.0.0?

Hmm.  I don't think there were any changes that would have effected this 
one way or
the other.

Cheers,
Leif