You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@cocoon.apache.org by Brent Johnson <bl...@gmail.com> on 2004/12/06 20:35:33 UTC

Site Stops Responding (2.1.5 w/ Tomcat 5.0.28)

I've brought this up on this list before (I think) - but its starting
to happen more frequently so I'm bringing it up again to see if anyone
has any ideas as to where I should spend my time investigating to see
what the problem is.

Cocoon version:  2.1.5
Tomcat Version:  5.0.28
OS:  Redhat Linux 9
Hardware:
Dell Poweredge 2600
Intel Xeon processor (2.4Ghz I think)
1GB Ram

I'm using the Tomcat HTTP connector to serve all pages.  I'm doing
this for a couple reasons... most notably ease.  I'm using cocoon
throughout the whole site as the base URI (i.e. www.someurl.com/).

After a few weeks (sometimes around a month.. this last time in about
3 weeks) the site just stops responding.  My browser tells me its
waiting for a response.. so a connection is being made.  After some
period of time.. bang.. nothing but a timeout and/or network error
from the browser.

This is on a production site.. luckily its a intranet site so I just
get calls from the employees trying to use it.  My fix is usually one
of the following:

1)  restart network script with /etc/rc.d/init.d/network restart
2)  if step #1 didnt work I shutdown and restart tomcat
3)  if neither of the two work (which has never happened so far) I'd
restart the machine

So what are my options here?  There seems to be no errors in the
tomcat logs and no errors in the cocoon logs (other than some piddly
handled errors about some missing stylesheet or something.. nothing
serious).

If ANYONE could give me a little insight.. or nudge me in the right
direction for getting this fixed I'd greatly appreciate it.  If it
wasnt for this problem.. the site would require no hours for
maintenance freeing me up for other work.

Thanks.

- Brent

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
For additional commands, e-mail: users-help@cocoon.apache.org


Re: Site Stops Responding (2.1.5 w/ Tomcat 5.0.28)

Posted by Bertrand Delacretaz <bd...@apache.org>.
Le 6 déc. 04, à 20:35, Brent Johnson a écrit :

> ....After a few weeks (sometimes around a month.. this last time in  
> about
> 3 weeks) the site just stops responding.  My browser tells me its
> waiting for a response.. so a connection is being made.  After some
> period of time.. bang.. nothing but a timeout and/or network error
> from the browser...

I'd start with by getting and analyzing a thread dump from the JVM.

There's some info at  
http://www.unixville.com/~moazam/stories/2004/05/18/ 
debuggingHangsInTheJvm.html for example, it's for Solaris but you can  
certainly find similar info for your platform.

-Bertrand

Re: Site Stops Responding (2.1.5 w/ Tomcat 5.0.28)

Posted by Jorg Heymans <jh...@domek.be>.
It's a bit tough debugging a problem when you only get a shot at it 
about once a month ;)

Brent Johnson wrote:
> I'll upgrade the system to Cocoon 2.1.6 when I get back from a
> business trip (cant use VPN thru the govmt network.. argh).  Hopefully
> this will take care of the problem.  If not I'll analyze the thread
> dump and send on the results (in about a month if it happens again).
> 
> Thanks for all the help.
> 
> - Brent
> 
> 
> On Tue, 07 Dec 2004 10:33:56 +0100, Jorg Heymans <jh...@domek.be> wrote:
> 
>>There was this "problem" [1] with semaphore handles that got "fixed" in
>>2.1.6.
>>
>>HTH
>>Jorg
>>
>>[1] http://issues.apache.org/bugzilla/show_bug.cgi?id=31760
>>
>>
>>
>>
>>Brent Johnson wrote:
>>
>>>I've brought this up on this list before (I think) - but its starting
>>>to happen more frequently so I'm bringing it up again to see if anyone
>>>has any ideas as to where I should spend my time investigating to see
>>>what the problem is.
>>>
>>>Cocoon version:  2.1.5
>>>Tomcat Version:  5.0.28
>>>OS:  Redhat Linux 9
>>>Hardware:
>>>Dell Poweredge 2600
>>>Intel Xeon processor (2.4Ghz I think)
>>>1GB Ram
>>>
>>>I'm using the Tomcat HTTP connector to serve all pages.  I'm doing
>>>this for a couple reasons... most notably ease.  I'm using cocoon
>>>throughout the whole site as the base URI (i.e. www.someurl.com/).
>>>
>>>After a few weeks (sometimes around a month.. this last time in about
>>>3 weeks) the site just stops responding.  My browser tells me its
>>>waiting for a response.. so a connection is being made.  After some
>>>period of time.. bang.. nothing but a timeout and/or network error
>>>from the browser.
>>>
>>>This is on a production site.. luckily its a intranet site so I just
>>>get calls from the employees trying to use it.  My fix is usually one
>>>of the following:
>>>
>>>1)  restart network script with /etc/rc.d/init.d/network restart
>>>2)  if step #1 didnt work I shutdown and restart tomcat
>>>3)  if neither of the two work (which has never happened so far) I'd
>>>restart the machine
>>>
>>>So what are my options here?  There seems to be no errors in the
>>>tomcat logs and no errors in the cocoon logs (other than some piddly
>>>handled errors about some missing stylesheet or something.. nothing
>>>serious).
>>>
>>>If ANYONE could give me a little insight.. or nudge me in the right
>>>direction for getting this fixed I'd greatly appreciate it.  If it
>>>wasnt for this problem.. the site would require no hours for
>>>maintenance freeing me up for other work.
>>>
>>>Thanks.
>>>
>>>- Brent
>>
>>---------------------------------------------------------------------
>>To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
>>For additional commands, e-mail: users-help@cocoon.apache.org
>>
>>


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
For additional commands, e-mail: users-help@cocoon.apache.org


Re: Site Stops Responding (2.1.5 w/ Tomcat 5.0.28)

Posted by Brent Johnson <bl...@gmail.com>.
I'll upgrade the system to Cocoon 2.1.6 when I get back from a
business trip (cant use VPN thru the govmt network.. argh).  Hopefully
this will take care of the problem.  If not I'll analyze the thread
dump and send on the results (in about a month if it happens again).

Thanks for all the help.

- Brent


On Tue, 07 Dec 2004 10:33:56 +0100, Jorg Heymans <jh...@domek.be> wrote:
> There was this "problem" [1] with semaphore handles that got "fixed" in
> 2.1.6.
> 
> HTH
> Jorg
> 
> [1] http://issues.apache.org/bugzilla/show_bug.cgi?id=31760
> 
> 
> 
> 
> Brent Johnson wrote:
> > I've brought this up on this list before (I think) - but its starting
> > to happen more frequently so I'm bringing it up again to see if anyone
> > has any ideas as to where I should spend my time investigating to see
> > what the problem is.
> >
> > Cocoon version:  2.1.5
> > Tomcat Version:  5.0.28
> > OS:  Redhat Linux 9
> > Hardware:
> > Dell Poweredge 2600
> > Intel Xeon processor (2.4Ghz I think)
> > 1GB Ram
> >
> > I'm using the Tomcat HTTP connector to serve all pages.  I'm doing
> > this for a couple reasons... most notably ease.  I'm using cocoon
> > throughout the whole site as the base URI (i.e. www.someurl.com/).
> >
> > After a few weeks (sometimes around a month.. this last time in about
> > 3 weeks) the site just stops responding.  My browser tells me its
> > waiting for a response.. so a connection is being made.  After some
> > period of time.. bang.. nothing but a timeout and/or network error
> > from the browser.
> >
> > This is on a production site.. luckily its a intranet site so I just
> > get calls from the employees trying to use it.  My fix is usually one
> > of the following:
> >
> > 1)  restart network script with /etc/rc.d/init.d/network restart
> > 2)  if step #1 didnt work I shutdown and restart tomcat
> > 3)  if neither of the two work (which has never happened so far) I'd
> > restart the machine
> >
> > So what are my options here?  There seems to be no errors in the
> > tomcat logs and no errors in the cocoon logs (other than some piddly
> > handled errors about some missing stylesheet or something.. nothing
> > serious).
> >
> > If ANYONE could give me a little insight.. or nudge me in the right
> > direction for getting this fixed I'd greatly appreciate it.  If it
> > wasnt for this problem.. the site would require no hours for
> > maintenance freeing me up for other work.
> >
> > Thanks.
> >
> > - Brent
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
> For additional commands, e-mail: users-help@cocoon.apache.org
> 
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
For additional commands, e-mail: users-help@cocoon.apache.org


Re: Site Stops Responding (2.1.5 w/ Tomcat 5.0.28)

Posted by Jorg Heymans <jh...@domek.be>.
There was this "problem" [1] with semaphore handles that got "fixed" in 
2.1.6.

HTH
Jorg

[1] http://issues.apache.org/bugzilla/show_bug.cgi?id=31760


Brent Johnson wrote:
> I've brought this up on this list before (I think) - but its starting
> to happen more frequently so I'm bringing it up again to see if anyone
> has any ideas as to where I should spend my time investigating to see
> what the problem is.
> 
> Cocoon version:  2.1.5
> Tomcat Version:  5.0.28
> OS:  Redhat Linux 9
> Hardware:
> Dell Poweredge 2600
> Intel Xeon processor (2.4Ghz I think)
> 1GB Ram
> 
> I'm using the Tomcat HTTP connector to serve all pages.  I'm doing
> this for a couple reasons... most notably ease.  I'm using cocoon
> throughout the whole site as the base URI (i.e. www.someurl.com/).
> 
> After a few weeks (sometimes around a month.. this last time in about
> 3 weeks) the site just stops responding.  My browser tells me its
> waiting for a response.. so a connection is being made.  After some
> period of time.. bang.. nothing but a timeout and/or network error
> from the browser.
> 
> This is on a production site.. luckily its a intranet site so I just
> get calls from the employees trying to use it.  My fix is usually one
> of the following:
> 
> 1)  restart network script with /etc/rc.d/init.d/network restart
> 2)  if step #1 didnt work I shutdown and restart tomcat
> 3)  if neither of the two work (which has never happened so far) I'd
> restart the machine
> 
> So what are my options here?  There seems to be no errors in the
> tomcat logs and no errors in the cocoon logs (other than some piddly
> handled errors about some missing stylesheet or something.. nothing
> serious).
> 
> If ANYONE could give me a little insight.. or nudge me in the right
> direction for getting this fixed I'd greatly appreciate it.  If it
> wasnt for this problem.. the site would require no hours for
> maintenance freeing me up for other work.
> 
> Thanks.
> 
> - Brent


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
For additional commands, e-mail: users-help@cocoon.apache.org


Re: Site Stops Responding (2.1.5 w/ Tomcat 5.0.28)

Posted by Brent Johnson <bl...@gmail.com>.
On Mon, 6 Dec 2004 11:44:28 -0800, Mike Dickson <mi...@yahoo.com> wrote:
> Could it be a memory leak? It sounds like it.

That was my initial thought.  The site uses XSP (no flow).  I'm
thinking about redeveloping the site using Flowscript and JXTemplates
instead of XSP.  But since its production I hate to mess with it too
much.

I did check the memory for the processes using the "ps" command (and
"top").. but I dunno if that would really indicate a memory leak.  The
tomcat process is still running and it appears to not be taking up any
real large amount of memory (if I recall correctly.. I dont think its
taking up any more when its failing than when its freshly started).

I'll check into it though.. oh.. one more thing.. the site doesnt get
a lot of usage so its pretty much idle most of the time.  But even
idle, it still takes about the same time (a month) for it to stop
responding.

On Mon, 6 Dec 2004 20:47:11 +0100, Bertrand Delacretaz
<bd...@apache.org> wrote:
> Le 6 déc. 04, à 20:35, Brent Johnson a écrit :
> 
> I'd start with by getting and analyzing a thread dump from the JVM.
> 
> There's some info at
> http://www.unixville.com/~moazam/stories/2004/05/18/
> debuggingHangsInTheJvm.html for example, it's for Solaris but you can
> certainly find similar info for your platform.

Thanks - I'll definitely give this a try when the process hangs again.
 Hopefully that will give me a better idea of whats going on.  It just
drives me crazy when something fails (esp. in production) but doesnt
tell me why its failing.

Thanks,

- Brent

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
For additional commands, e-mail: users-help@cocoon.apache.org