Re: [Wrapper-user] ^^C not working sometimes

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

Leif,

Here is the log during the double ctl-c.

It takes 6 seconds to shutdown the JVM starting at

DEBUG  | wrapper  | main    | 2007/11/20 13:58:36 | Signal trapped.  Detail=
s:
DEBUG  | wrapper  | main    | 2007/11/20 13:58:36 |   signal number=3D2 (SI=
GINT), source=3D"the kernel"
STATUS | wrapper  | main    | 2007/11/20 13:58:36 | INT trapped.  Shutting =
down.
DEBUG  | wrapper  | main    | 2007/11/20 13:58:36 | wrapperStopProcess(0) c=
alled.
DEBUG  | wrapper  | main    | 2007/11/20 13:58:36 | Sending stop signal to =
JVM
DEBUG  | wrapperp | main    | 2007/11/20 13:58:36 | send a packet STOP : NU=
LL
DEBUG  | wrapper  | main    | 2007/11/20 13:58:36 | Signal trapped.  Detail=
s:
DEBUG  | wrapper  | main    | 2007/11/20 13:58:36 |   signal number=3D2 (SI=
GINT), source=3D"the kernel"
STATUS | wrapper  | main    | 2007/11/20 13:58:36 | INT trapped.  Forcing i=
mmediate shutdown.
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: R=
eceived a packet STOP :
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: T=
hread, Wrapper-Connection, handling the shutdown process.
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: c=
alling listener.stop()
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: W=
aiting for WrapperListener.stop runner thread to complete.
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: P=
rocessing control event(WRAPPER_CTRL_C_EVENT)
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: W=
rapperManager.stop(0) called by thread: Wrapper-Control-Event-Monitor
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: T=
hread, Wrapper-Control-Event-Monitor, waiting for the JVM to exit.
INFO   | jvm 1    | main    | 2007/11/20 13:58:36 | WrapperManager Debug: W=
rapperListener.stop runner thread started.

< OUR JVM SHUTDOWN PROCESSING HERE>
.
.
.
< OUR JVM SHUTDOWN PROCESSING HERE>

and ending at:

INFO   | jvm 1    | main    | 2007/11/20 13:58:42 | WrapperManager Debug: W=
rapperListener.stop runner thread stopped.
INFO   | jvm 1    | main    | 2007/11/20 13:58:42 | WrapperManager Debug: r=
eturned from listener.stop() -> 0
INFO   | jvm 1    | main    | 2007/11/20 13:58:42 | WrapperManager Debug: s=
hutdownJVM(0) Thread:Wrapper-Connection
INFO   | jvm 1    | main    | 2007/11/20 13:58:42 | WrapperManager Debug: S=
end a packet STOPPED : 0
INFO   | jvm 1    | main    | 2007/11/20 13:58:42 | WrapperManager Debug: C=
losing socket.
INFO   | jvm 1    | main    | 2007/11/20 13:58:42 | WrapperManager Debug: c=
alling System.exit(0)
DEBUG  | wrapper  | main    | 2007/11/20 13:58:42 | JVM exited normally.
DEBUG  | wrapper  | main    | 2007/11/20 13:58:42 | Signal trapped.  Detail=
s:
DEBUG  | wrapper  | main    | 2007/11/20 13:58:42 |   signal number=3D17 (S=
IGCHLD), source=3D"unknown"
DEBUG  | wrapper  | main    | 2007/11/20 13:58:42 | Received SIGCHLD, check=
ing JVM process status.
DEBUG  | wrapper  | main    | 2007/11/20 13:58:42 | JVM process exited with=
 a code of 0, leaving the wrapper exit code set to 0.
STATUS | wrapper  | main    | 2007/11/20 13:58:42 | <-- Wrapper Stopped

It seems like that even when the wrapper says it is forcing an immediate sh=
utdown, it is waiting for the JVM to finish and during that time our code i=
s doing its shutdown processing, which takes 6 seconds in this case.

I'm confused regarding the notion of "immediate".  What is an immediate shu=
tdown?  How is it different than kill -9 JVM_PID?

With the previous wrapper (2.3) this never happened.  Did anything change t=
hat would affect immediate shutdown?

Thanks for your help,

Tom

-----Original Message-----
From: wra...@li... [mailto:wrapper-user-bounc=
es...@li...] On Behalf Of Leif Mortenson
Sent: Monday, November 19, 2007 6:28 PM
To: wra...@li...
Subject: Re: [Wrapper-user] ^^C not working sometimes

Tom,
There used to be a bug where the Wrapper would sometimes mistake
a single CTRL-C event as two and kill the JVM.  But I have never seen
what you are seeing.

Could you try to reproduce this with wrapper.debug=3Dtrue enabled?

How long apart are the two CTRL-C events?   Are you seeing
messages in the log for both of them?

Cheers,
Leif

Tom Saulpaugh wrote:
>
> Has anyone noticed that double cntrl-c  doesn't always kill the JVM?
>
> I have seen this on linux and windows.
>
> Is there a fix for this?
>
> It seems to happen when the JVM is under heavy load.
>
>
>
> Thanks,
>
>
>
> Tom
>
>
>
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> ------------------------------------------------------------------------
>
> _______________________________________________
> Wrapper-user mailing list
> Wra...@li...
> https://lists.sourceforge.net/lists/listinfo/wrapper-user
>

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Wrapper-user mailing list
Wra...@li...
https://lists.sourceforge.net/lists/listinfo/wrapper-user