Hello, Interesting, MaxRequestsPerChild limit is to low. How I should set the limit? I have made a downgrade to mod_jk 1.2.27
Some additionally logs from error.log: [Mon Feb 01 09:45:00 2010] [notice] Parent: Received shutdown signal -- Shutting down the server. [Mon Feb 01 09:45:00 2010] [notice] Child 9668: Exit event signaled. Child process is ending. [Mon Feb 01 09:45:00 2010] [warn] (OS 10038)An operation was attempted on something that is not a socket. : setsockopt(SO_UPDATE_ACCEPT_CONTEXT) failed. [Mon Feb 01 09:45:01 2010] [notice] Child 9668: Released the start mutex [Mon Feb 01 09:45:30 2010] [notice] Parent: Forcing termination of child process 440 [Mon Feb 01 09:45:36 2010] [notice] Apache/2.2.14 (Win32) mod_jk/1.2.28 configured -- resuming normal operations [Mon Feb 01 09:45:36 2010] [notice] Server built: Sep 28 2009 22:41:08 [Mon Feb 01 09:45:36 2010] [notice] Parent: Created child process 10940 [Mon Feb 01 09:45:37 2010] [notice] Child 10940: Child process is running [Mon Feb 01 09:45:37 2010] [notice] Child 10940: Acquired the start mutex. [Mon Feb 01 09:45:37 2010] [notice] Child 10940: Starting 1920 worker threads. [Mon Feb 01 09:45:38 2010] [notice] Child 10940: Starting thread to listen on port 80. [Mon Feb 01 10:32:53 2010] [notice] Child 10940: Process exiting because it reached MaxRequestsPerChild. Signaling the parent to restart a new child process. [Mon Feb 01 10:32:53 2010] [notice] Parent: Received restart signal -- Restarting the server. [Mon Feb 01 10:32:53 2010] [warn] (OS 10038)An operation was attempted on something that is not a socket. : setsockopt(SO_UPDATE_ACCEPT_CONTEXT) failed. [Mon Feb 01 10:32:54 2010] [notice] Apache/2.2.14 (Win32) mod_jk/1.2.28 configured -- resuming normal operations [Mon Feb 01 10:32:54 2010] [notice] Server built: Sep 28 2009 22:41:08 [Mon Feb 01 10:32:54 2010] [notice] Parent: Created child process 2032 [Mon Feb 01 10:32:54 2010] [notice] Child 10940: Released the start mutex [Mon Feb 01 10:32:55 2010] [notice] Child 2032: Child process is running [Mon Feb 01 10:32:55 2010] [notice] Child 2032: Acquired the start mutex. [Mon Feb 01 10:32:55 2010] [notice] Child 2032: Starting 1920 worker threads. [Mon Feb 01 10:32:56 2010] [notice] Child 2032: Starting thread to listen on port 80. [Mon Feb 01 10:33:01 2010] [notice] Child 10940: All worker threads have exited. [Mon Feb 01 10:33:01 2010] [notice] Child 10940: Child process is exiting [Mon Feb 01 11:19:17 2010] [notice] Child 2032: Process exiting because it reached MaxRequestsPerChild. Signaling the parent to restart a new child process. [Mon Feb 01 11:19:17 2010] [notice] Parent: Received restart signal -- Restarting the server. [Mon Feb 01 11:19:17 2010] [warn] (OS 10038)An operation was attempted on something that is not a socket. : setsockopt(SO_UPDATE_ACCEPT_CONTEXT) failed. [Mon Feb 01 11:19:18 2010] [notice] Apache/2.2.14 (Win32) mod_jk/1.2.28 configured -- resuming normal operations [Mon Feb 01 11:19:18 2010] [notice] Server built: Sep 28 2009 22:41:08 [Mon Feb 01 11:19:18 2010] [notice] Parent: Created child process 4892 [Mon Feb 01 11:19:18 2010] [notice] Child 2032: Released the start mutex [Mon Feb 01 11:19:19 2010] [notice] Child 4892: Child process is running [Mon Feb 01 11:19:19 2010] [notice] Child 4892: Acquired the start mutex. [Mon Feb 01 11:19:19 2010] [notice] Child 4892: Starting 1920 worker threads. [Mon Feb 01 11:19:19 2010] [notice] Child 4892: Starting thread to listen on port 80. [Mon Feb 01 11:19:53 2010] [notice] Child 2032: Waiting 270 more seconds for 1 worker threads to finish. [Mon Feb 01 11:20:02 2010] [notice] Parent: Received shutdown signal -- Shutting down the server. [Mon Feb 01 11:20:02 2010] [notice] Child 4892: Exit event signaled. Child process is ending. [Mon Feb 01 11:20:03 2010] [notice] Child 4892: Released the start mutex [Mon Feb 01 11:20:04 2010] [notice] Child 4892: Terminating 1035 threads that failed to exit. [Mon Feb 01 11:20:04 2010] [notice] Child 4892: All worker threads have exited. [Mon Feb 01 11:20:04 2010] [notice] Child 4892: Child process is exiting [Mon Feb 01 11:20:04 2010] [notice] Parent: Child process exited successfully. [Mon Feb 01 11:20:26 2010] [notice] Child 2032: Waiting 240 more seconds for 1 worker threads to finish. [Mon Feb 01 11:20:27 2010] [notice] Apache/2.2.14 (Win32) mod_jk/1.2.27 configured -- resuming normal operations [Mon Feb 01 11:20:27 2010] [notice] Server built: Sep 28 2009 22:41:08 [Mon Feb 01 11:20:27 2010] [notice] Parent: Created child process 2112 [Mon Feb 01 11:20:28 2010] [notice] Child 2112: Child process is running [Mon Feb 01 11:20:28 2010] [notice] Child 2112: Acquired the start mutex. [Mon Feb 01 11:20:28 2010] [notice] Child 2112: Starting 1920 worker threads. [Mon Feb 01 11:20:29 2010] [notice] Child 2112: Starting thread to listen on port 80. [Mon Feb 01 11:21:00 2010] [notice] Child 2032: Waiting 210 more seconds for 1 worker threads to finish. -----Ursprüngliche Nachricht----- Von: Alexander Diedler [mailto:adied...@tecracer.de] Gesendet: Montag, 1. Februar 2010 08:40 An: Tomcat Users List Betreff: AW: Tomcat Cluster 6.x and HTTP/1.1 503 Service Temporarily Unavailable Hello, Server Version: Apache/2.2.14 (Win32) mod_jk/1.2.28 Server Built: Sep 28 2009 22:41:08 Server loaded APR Version: 1.3.9 Compiled with APR Version: 1.3.9 Server loaded APU Version: 1.3.9 Compiled with APU Version: 1.3.9 Module Magic Number: 20051115:23 Hostname/port: XXX.XXX.XXX.XXX:80 Timeouts: connection: 300 keep-alive: 5 MPM Name: WinNT MPM Information: Max Daemons: 64 Threaded: yes Forked: no Server Architecture: 32-bit Server Root: D:/tecracer/software/Apache Config File: D:/tecracer/software/Apache/conf/httpd.conf Server Built With: -D APACHE_MPM_DIR="server/mpm/winnt" -D APR_HAS_SENDFILE -D APR_HAS_MMAP -D APR_HAS_OTHER_CHILD -D AP_HAVE_RELIABLE_PIPED_LOGS -D HTTPD_ROOT="/apache" -D SUEXEC_BIN="/apache/bin/suexec" -D DEFAULT_ERRORLOG="logs/error.log" -D AP_TYPES_CONFIG_FILE="conf/mime.types" -D SERVER_CONFIG_FILE="conf/httpd.conf" Server Version: Apache/2.2.14 (Win32) mod_jk/1.2.28 Server Built: Sep 28 2009 22:41:08 -------------------------------------------------------------------------------- Current Time: Monday, 01-Feb-2010 08:38:14 W. Europe Standard Time Restart Time: Monday, 01-Feb-2010 04:58:09 W. Europe Standard Time Parent Server Generation: 0 Server uptime: 3 hours 40 minutes 5 seconds Total accesses: 28126 - Total Traffic: 340.1 MB 2.13 requests/sec - 26.4 kB/second - 12.4 kB/request 44 requests currently being processed, 1656 idle workers Listing Load Balancing Worker (1 Worker) [Hide] -------------------------------------------------------------------------------- [S|E|R] Worker Status for loadbalancer Type Sticky Sessions Force Sticky Sessions Retries LB Method Locking Recover Wait Time Error Escalation Time Max Reply Timeouts [Hide] lb True False 3 Busyness Optimistic 60 30 0 Good Degraded Bad/Stopped Busy Max Busy Next Maintenance Last Reset [Hide] 2 0 0 0 70 74/196 13277 Balancer Members [Hide] Name Type Hostname Address:Port Connection Pool Timeout Connect Timeout Prepost Timeout Reply Timeout Retries Recovery Options Max Packet Size [Hide] worker1 ajp13 k0.domain.local xxx.xxx.x.78:8010 0 0 0 0 2 0 8192 worker2 ajp13 k1.domain.local xxx.xxx.x.77:8010 0 0 0 0 2 0 8192 Name Act State D F M V Acc Err CE RE Wr Rd Busy Max Route RR Cd Rs LR LE [S|E|R] worker1 ACT OK 0 100 1 0 425 (1/sec) 0 0 0 264K (908 /sec) 1.6M (5.3K/sec) 0 5 worker1 0/0 298 [S|E|R] worker2 ACT OK 0 100 1 0 2257 (0/sec) 112 24 0 1.3M (104 /sec) 17M (1.3K/sec) 0 11 worker2 0/0 13277 Greetings Alexander Diedler -----Ursprüngliche Nachricht----- Von: Mark Thomas [mailto:ma...@apache.org] Gesendet: Sonntag, 31. Januar 2010 18:55 An: Tomcat Users List Betreff: Re: Tomcat Cluster 6.x and HTTP/1.1 503 Service Temporarily Unavailable On 31/01/2010 17:50, Alexander Diedler wrote: > Hello, > We spend a lot of time to investigate the mistake, but I would not be found > by us. We have two Virtual Machines with Windows 2003 R2 Standard Edition and > on every node there is a Apache 2.2.14 and Tomcat 6.0.20 (and tcnative > installed) with mod_jk 1.2.28 (all 32 Bit) installed. > After a period (hours up to a day) we receive on every application a 503 > error through the Webserver. > The Apache seems to work fine, jkmanager, htdocs, server-info and > server-status can be opened and the apps seems can be accessed by HTTP > Connector Port 8080. > > Is there are any idea, what is the problem in this case? Any suggestions, > what we can improve to avoid this errors. Hard to tell with out any configuration info. At a guess, more httpd threads than AJP threads and unused connections not being closed quickly enough. Try fixing your timeouts or use JkOptions +DisableReuse Mark --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org