You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@tomcat.apache.org by Marek Lange <ml...@objectone.biz> on 2004/06/04 17:29:02 UTC

mod_jk2 connection problems

I have set up an enivronment to test mod_jk2 with JBoss 3.2.3 (embedded 
Tomcat 4.1.x). I compiled mod_jk2 (2.0.4) for Linux (2.6) by myself. 
Everything seems to be working so far. Requests from Apache (2.0.49) are 
forwarded to the tomcat instances.

The problem occurs when I try to connect after tomcat has fired the 
message "[ChannelSocket] connection timeout reached". Apache is now 
giving the error "Service Temporarily Unavailable The server is 
temporarily unable to service your request due to maintenance downtime 
or capacity problems. Please try again later."). HTTP 503. The mod_jk2 
log says:

[Fri Jun 04 15:45:26 2004] (debug ) [jk_worker_lb.c (399)]  lb.service() 
try ajp13:instance2:7011
[Fri Jun 04 15:45:26 2004] (error ) [jk_channel_apr_socket.c (405)] 
jk2_channel_apr_send send failed 32 Broken pipe
[Fri Jun 04 15:45:26 2004] (error ) [jk_worker_ajp13.c (277)] 
ajp13.checkalive() can't send cping request to ajp13:instance2:7011
[Fri Jun 04 15:45:26 2004] (debug ) [jk_worker_ajp13.c (701)] 
ajp13.done() return to pool ajp13:instance2:7011
[Fri Jun 04 15:45:26 2004] (error ) [jk_worker_lb.c (427)]  lb.service() 
worker failed 120000 for ajp13:instance2:7011
[Fri Jun 04 15:45:26 2004] (debug ) [jk_worker_lb.c (451)] 
lb_worker.service() try other hosts
[Fri Jun 04 15:45:26 2004] (debug ) [jk_worker_lb.c (88)]  lb.get_worker 
1 Session=NULL
[Fri Jun 04 15:45:26 2004] ( info ) [jk_worker_lb.c (227)] 
lb.getWorker() All workers in error state, use the one with oldest error
[Fri Jun 04 15:45:26 2004] ( info ) [jk_worker_lb.c (262)] 
lb.getWorker() We tried all possible workers 2
[Fri Jun 04 15:45:26 2004] (error ) [jk_worker_lb.c (367)] 
lb_worker.service() all workers in error or disabled state
[Fri Jun 04 15:45:26 2004] (error ) [mod_jk2.c (755)]  mod_jk2.handler() 
Error connecting to tomcat 120000, status 503

jk_status states that all the workers are now in "errorState". Tomcat is 
still answering well when asked via the HTTP connector, not the AJP1.3 
connector. My questions:

Is that a configuration problem of mod_jk2?
How can I set the workers out of the error state?
Why would jk2.properties be needed in Tomcat conf (are defaults for this 
ok)?
What about the stability of mod_jk2? Can it be used in production with 
high load factors? Experiences?

Thanks!

-marek

---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-user-help@jakarta.apache.org