I have read similar thread at jonas mailing list. The problem is that we are
using CMP.
Thanks,
Miro
-----Original Message-----
From: Eric Richardson [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, March 21, 2000 4:56 PM
To: Halas, Miroslav
Cc: '[EMAIL PROTECTED]'; '[EMAIL PROTECTED]'
Subject: Re: Connection manager troubles...
Hi,
I ran into this underflow with BMP where I either forgot to close the
connection or closed
it twice. When we matched getConnection() with close() the error went away.
Don't know
if that will solve your problem but it could help.
Eric :-)
"Halas, Miroslav" wrote:
> Hello,
>
> We are running Jonas 1.6.1 on Jeremie with Interbase 5.5 and Interclient
> 1.51
> Our server sometimes just hang up and the processor is on 100% utilization
> even
> if nobody is doing anything. Last time this happened, I noticed following
> output, which I assume came from Jonas
>
> WARNING: Connection was not always closed correctly !
> connection-open counter underflow
> removeConnection: connectionlist is empty
>
> Does it ring a bell? Have anybody seen something similar? ANy help is
> greatly appreciated.
>
> Regards,
>
> Miroslav Halas
> Software Engineer
> Compuware Corp.
> 15305 Dallas Parkway
> Suite 900
> Addison, TX 75001
> phone 9720-960-0960 x 1333
> fax 972-960-8489
>
> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, February 29, 2000 9:08 AM
> To: Eric Richardson; [EMAIL PROTECTED]
> Subject: Re: Connection manager troubles...
>
> Yeah... It turns out that this may have cropped up because we moved all
our
> read only methods over to tx_supports... now the connection grabbing is
> happening outside a transaction, so it is relying on us to be good
> consistent
> programmers, instead of letting us be sloppy and having the end trans
clean
> everything up. I suspect in most places we are clean, but we must have
> missed
> one.... I'm still a little skeptical though because it seems that if we
run
> through the same set of bean calls, we dont see a problem til we have
> several
> threads banging on it for around two hours... I'd expect to see something
> right away if it was our code forgetting to close something....
>
> I'm not ready to give up and say its our code yet because of this, our
> current
> theory is that it may be a multithreaded issue... but I'm also trying to
> chase
> down if we are fogetting to close a connection in our code somewhere or
> closing it more then once....
>
> -gabe
>
> Eric Richardson wrote:
>
> > Hi,
> > We were using Bean Managed Persistance and ran into an underflow
problem.
> > We were
> > calling an extra close on the connection. It been awhile so my memory is
a
> > bit fuzzy. It was
> > after the connection pool was revamped however.
> >
> > Hopefully, this will give you a starting place.
> > Eric :-)
> ----
> To unsubscribe, send email to [EMAIL PROTECTED] and
> include in the body of the message "unsubscribe jonas-users".
> For general help, send email to [EMAIL PROTECTED] and
> include in the body of the message "help".
----
This list is cross-posted to two mail lists. To unsubscribe,
follow the instructions below for the list you subscribed to.
For objectweb.org: send email to [EMAIL PROTECTED] and
include in the body of the message "unsubscribe dbtm-group".
For enhydra.org: send email to [EMAIL PROTECTED] and include
in the body of the message "unsubscribe dbtm-group".
