Sorry, Adrian - out of ideas, and Googling around for the error messages doesn't help much.
I would certainly try updating mod_jk to the latest version, or recompiling from the version shipped with whichever Tomcat distribution you're using, all the while keeping your fingers crossed! However, it does feel more like a problem at the Tomcat end. If you are completely free to use whichever version of Tomcat you like, and you're already running web apps using 5.5.x, then you could try upgrading to 5.5.15, recently out of beta. Good luck - Jon -----Original Message----- From: Adrian Nadeau [mailto:[EMAIL PROTECTED] Sent: 26 January 2006 17:01 To: Jonathan Woods Cc: 'Tomcat Users List' Subject: Re: Tomcat and mod_jk question on Red Hat 9 Thanks for the information... In our mod_jk log we have the following showing up: [Wed Jan 25 13:21:33 2006] [jk_ajp_common.c (1401)]: ERROR: Receiving from tomcat failed, recoverable operation. err=1 [Wed Jan 25 13:21:33 2006] [jk_ajp_common.c (836)]: ajp_connection_tcp_get_message: Error - Wrong message format 0x4854 [Wed Jan 25 13:21:33 2006] [jk_ajp_common.c (1248)]: Error reading reply from tomcat. Tomcat is down or network problems. [Wed Jan 25 13:21:33 2006] [jk_ajp_common.c (1401)]: ERROR: Receiving from tomcat failed, recoverable operation. err=2 [Wed Jan 25 13:21:33 2006] [jk_ajp_common.c (1429)]: Error connecting to tomcat. Tomcat is probably not started or is listening on the wrong port. worker=ajp13 failed errno = 11 That's after it actually works a few times, so Tomcat is definitely running and accessible (thru port 8080 also). In error log there a lot of processes that haven't exited, here is some information from it when we actually do a restart: start[Thu Jan 26 10:31:24 2006] [warn] child process 5805 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 5806 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 6668 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 28585 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 13921 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 27065 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 13928 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 27152 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 18708 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:24 2006] [warn] child process 18714 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 5805 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 5806 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 6668 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 28585 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 13921 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 27065 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 13928 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 27152 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 18708 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:25 2006] [warn] child process 18714 still did not exit, sending a SIGTERM [Thu Jan 26 10:31:26 2006] [notice] caught SIGTERM, shutting down [Thu Jan 26 10:31:31 2006] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec) [Thu Jan 26 10:31:31 2006] [notice] Digest: generating secret for digest authentication ... [Thu Jan 26 10:31:31 2006] [notice] Digest: done [Thu Jan 26 10:31:32 2006] [notice] Apache/2.0.40 (Red Hat Linux) configured -- resuming normal operations I don't believe it's our browser, we can actually reopen a new browser and it won't work right also (if we do it around the same time). I'm going to check the mod_jk file again and make 100% sure it's compatable with our versions of Tomcat and Apache. Do you think this could be causing the problem possibly? Any other suggestions would be great, Thanks again. Adrian Nadeau VP, Development Evolving Solutions...Technology for changing [EMAIL PROTECTED] www.evolvingsolutions.ca 506.633.2012 Jonathan Woods wrote: >Adrian - > >I haven't met this problem, but a few things occurred to me after >reading your message: > >1. Don't forgot you could probably (depending on your config) always >use mod_proxy instead of mod_jk, just as a temporary measure. > >2. Sometimes browsers are set to have only 2 or 3 concurrent >connections open to a server. Maybe something about your config - e.g. >the resources being served up, or a mod_jk problem - means that the >connections aren't properly closed, so on the third click it's actually >your browser waiting until the other connections are closed. > >3. It also sounds like a threading problem could cause these symptoms. >Is there any 2- or 3-ness about the Tomcat config - e.g. the size of >the pool of servlet instances maintained? Are the resources you're >reaching when you click threadsafe? > >4. Try looking in Apache's error log and in mod_jk's log (assuming you >write one - JkLogFile directive). > >Jon > >-----Original Message----- >From: Adrian Nadeau [mailto:[EMAIL PROTECTED] >Sent: 26 January 2006 14:56 >To: users@tomcat.apache.org >Subject: Tomcat and mod_jk question on Red Hat 9 > >Hello, > >We recently setup mod_jk on a server running Red Hat 9. We are having >some odd problems with the new setup. For some reason, everything >works fine for the first 2 clicks when testing the process (running the >examples webapp thru Apache 2.0.40). However, when we click something >(anything, doesn't have to be the same steps) for the third time, it >seems as though mod_jk cannot connect through AJP13 (Tomcat version is >4.1.29) . It trys to load for a long time and then nothing loads. If >we wait for a bit and then do the same test again, the same thing happens. > >I can post more information and config settings, but I wanted to see if >this is a common problem or not? We've had no problems running mod_jk >in a Windows environment or Solaris in the past. If we need to post >more information that is not a problem, we can do that. We talked to >our hosting provider and they seem to think it's that we are running out of TCP Sockets? >Any information would be great as we are trying to get a number of >applications running on the server ASAP (of course). Thanks! > >Regards, > >Adrian Nadeau >VP, Development >Evolving Solutions...Technology for changing >[EMAIL PROTECTED] www.evolvingsolutions.ca >506.633.2012 > >--------------------------------------------------------------------- >To unsubscribe, e-mail: [EMAIL PROTECTED] >For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
