Re: Choosing the right worker
Hi, We are not observing a failover. Our index page served by our first app server (worker1) was returning a 500 http status code, whereas the same index page served by our second app server (worker2) was returning normally (http status = 200). Yet the 500 error page was getting all the way back to the client. The docs say, The redirect flag on worker1 tells the lb_worker to redirect the requests to worker2 only if worker1 is in error state. In other cases worker2 will not receive any requests, thus acting like a hot standby. But we are still seeing our second app server get an equal number of requests. So could you clarify how to set up so that both nodes get an equal number of requests, but if one returns an error code, all requests will go to the other? Thanks, - Dave Christopher Schultz-2 wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Dave, On 2/23/2011 3:32 PM, laredotornado wrote: We're using 6.0.24, running on two servers. In our workers.properties file on the web server, we have ... ===Begin workers.properties == worker.lbroute.balance_workers=worker1,worker2 worker.worker1.type=ajp13 worker.worker1.host=ip_addr1 worker.worker1.port=8000 worker.worker1.lbfactor=1 You might want to look into using worker templates: you can set all of this stuff for a single template worker and inherit them in the individual ones. Useful for a consistent configuration for all workers in a lb cluster. worker.worker1.redirect=worker2 This should failover to worker2 if worker1 suffers an error, even though worker2 is set to disabled. The docs say as much: you can use this as a hot standby. You might want to look at the ping_mode, good, bad settings, too. How can I configure my workers.properties file so that if the index page of our sites is throwing a 500 http status code, it will fail over to the other worker? If your index page is being served by Tomcat/worker1, then this configuration should fail-over to worker2. Is that not what you are observing? If not, then please tell us what you are observing. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk1lkCsACgkQ9CaO5/Lv0PADywCfewAFL3/W5C7EVSL91NFxbBfA hNsAn34/DRp5KmNHJwb4P/b6co0bTRhq =pmf7 -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org -- View this message in context: http://old.nabble.com/Choosing-the-right-worker-tp30998571p31004353.html Sent from the Tomcat - User mailing list archive at Nabble.com. - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Choosing the right worker
Hi, We're using 6.0.24, running on two servers. In our workers.properties file on the web server, we have ... ===Begin workers.properties == worker.lbroute.balance_workers=worker1,worker2 worker.worker1.type=ajp13 worker.worker1.host=ip_addr1 worker.worker1.port=8000 worker.worker1.lbfactor=1 worker.worker1.redirect=worker2 worker.worker1.socket_keepalive=true worker.worker1.connection_pool_timeout=60 worker.worker2.type=ajp13 worker.worker2.host=ip_addr2 worker.worker2.port=8000 worker.worker2.lbfactor=1 worker.worker2.activation=disabled worker.worker2.socket_keepalive=true worker.worker2.connection_pool_timeout=60 ===End workers.properties == How can I configure my workers.properties file so that if the index page of our sites is throwing a 500 http status code, it will fail over to the other worker? - Dave -- View this message in context: http://old.nabble.com/Choosing-the-right-worker-tp30998571p30998571.html Sent from the Tomcat - User mailing list archive at Nabble.com. - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Choosing the right worker
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Dave, On 2/23/2011 3:32 PM, laredotornado wrote: We're using 6.0.24, running on two servers. In our workers.properties file on the web server, we have ... ===Begin workers.properties == worker.lbroute.balance_workers=worker1,worker2 worker.worker1.type=ajp13 worker.worker1.host=ip_addr1 worker.worker1.port=8000 worker.worker1.lbfactor=1 You might want to look into using worker templates: you can set all of this stuff for a single template worker and inherit them in the individual ones. Useful for a consistent configuration for all workers in a lb cluster. worker.worker1.redirect=worker2 This should failover to worker2 if worker1 suffers an error, even though worker2 is set to disabled. The docs say as much: you can use this as a hot standby. You might want to look at the ping_mode, good, bad settings, too. How can I configure my workers.properties file so that if the index page of our sites is throwing a 500 http status code, it will fail over to the other worker? If your index page is being served by Tomcat/worker1, then this configuration should fail-over to worker2. Is that not what you are observing? If not, then please tell us what you are observing. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk1lkCsACgkQ9CaO5/Lv0PADywCfewAFL3/W5C7EVSL91NFxbBfA hNsAn34/DRp5KmNHJwb4P/b6co0bTRhq =pmf7 -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org