Can you try the CVS head, and see if it still happens ?
If so, can you give us more details that would allow to
reproduce it ? 

Messages like 'it doesn't work for me, you figure out why'
are not very helpfull.

At least enable the debug messages, let it run until it
hungs, then maybe trim the log and send the last entries.
Or use gdb to attach to the problem process and do a 
stack trace. 

Costin

On Wed, 1 May 2002, Costas Stergiou wrote:

> Hi,
> I have an installation of Apache 1.3.19 + Tomcat 4.0.3
> on a aix4.3 machine with IBM JDK 1.3.0. I have two tomcat
> engines and I use the lb worker of mod_jk to balance user
> request to these two engines. 
> 
> The problem is that some httpd processes seem to go
> into some cpu performance tasks and start consuming all
> the cpu of the machine (2processors, 2GB mem). After
> some time almost all available CPU is consumed by the
> apache processes. If I kill the child processes that cause
> the problem, the load falls back to normal. At the same time,
> I notice that I have excessive CPU usage in one of the
> tomcat instances.
> 
> The problem keeps recurring all the time (almost every hour).
> I read in the mail-archives about similar problems but I didn't 
> find a good answer. 
> 
> The version of mod_jk is the one in jakarta-tomcat-connectors-4.0.2.01
> I compiled the source on the same machine (aix).
> 
> I have to say that I have noticed similar problems on many more 
> apache installations that have similar installations (apache, lb + 
> 2 or more tomcat engines) with older versions of mod_jk and tomcat.
> 
> Actually, the above problem makes the lb worker unusable since 
> I cannot trust the setup in my production environment. 
> 
> Please, any idea or patches or whatever will be very helpful.
> Thanks very much for any ideas,
> Costas
> 
> 
> 
> 


--
To unsubscribe:   <mailto:[EMAIL PROTECTED]>
For additional commands: <mailto:[EMAIL PROTECTED]>
Troubles with the list: <mailto:[EMAIL PROTECTED]>

Reply via email to