Re: [OT] RE: 24X7 deployment tips
I guess there is. By fixing the code. On Tue, 11 Jan 2005 18:05:42 +, Mark Benussi [EMAIL PROTECTED] wrote: The reason this is off topic is that I was wondering if anyone has ever found a way of recovering leaked memory, without restarting the JVM? - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Benefit of multiple workers?
I'm using mod_jk 1.2.5 with Apache2. The benefit is to balance the load between multiple app servers in a round robin manner, with a certain factor you can set for each worker with say, for example have different hardware specs. Regards, Faisal On Thu, 13 Jan 2005 18:24:27 -0500, John Martyniak [EMAIL PROTECTED] wrote: I am preparing to set up my first production Tomcat using mod_jk. So what is the benefit of multiple workers? I can see that you can specify one worker per host, so the advantage would be that you could have one Webserver drive several app servers. I also read on this list, that you could use mod_jk2 to do load balancing. (is the same true for mod_jk 1.2.x?) Are these the only benefits? Or am I completely missing the point? -John - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] -- There are many ways of going forward, but there is only one way of standing still. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Tomcat monitoring scripts
On demand restarts with: http://wrapper.tanukisoftware.org On Wed, 12 Jan 2005 16:50:54 +, Didier McGillis [EMAIL PROTECTED] wrote: btw the script I would use as a starting point, it doesnt work quite right on my system but I was looking for a starting point. From: Didier McGillis [EMAIL PROTECTED] Reply-To: Tomcat Users List tomcat-user@jakarta.apache.org To: [EMAIL PROTECTED], tomcat-user@jakarta.apache.org Subject: Re: Tomcat monitoring scripts Date: Wed, 12 Jan 2005 16:22:52 + Ch-Check this out. Shell script http://www.wespoke.com/archives/000728.php From: Edd Dawson [EMAIL PROTECTED] Reply-To: [EMAIL PROTECTED] To: Tomcat Users List tomcat-user@jakarta.apache.org Subject: Re: Tomcat monitoring scripts Date: Wed, 12 Jan 2005 15:20:53 + I'm looking more for something that sits on the actual servers and probes at set intervals and takes remedial action if necessary.. i just wondered if anyone had documented doing such a thing before. Peter Lin wrote: JMeter has a monitor for tomcat 5.0.19 and newer. It doesn't work with tomcat4 or older. in terms of restarting, you're probably going to have to write a shell script to do that. Typically, on unix a cron job is used. peter On Wed, 12 Jan 2005 15:12:15 +, Edd Dawson [EMAIL PROTECTED] wrote: Hi Does anyone know of the location of any scripts (for Linux) that will monitor tomcat every x minutes and if it finds it not running will restart it automatically? Any suggestions would be appreciated. thanks Edd - 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] -- Edd Dawson Head of Development MitchellConnerSearson 3-5 High Pavement The Lace Market Nottingham NG1 1HF Tel +44 (0)115 959 6455 Fax +44 (0)115 959 6456 Direct +44 (0)115 959 6463 www.choosemcs.co.uk http://www.choosemcs.co.uk/ Confidentiality: This e-mail and its attachments are intended for the above named only and may be confidential. If they have come to you in error you must take no action based on them, nor must you copy or show them to anyone; please reply to this e-mail and highlight the error. Security Warning: Please note that this e-mail has been created in the knowledge that Internet e-mail is not a 100% secure communications medium. We advise that you understand and observe this lack of security when e-mailing us. Viruses: Although we have taken steps to ensure that this e-mail and attachments are free from any virus, we advise that in keeping with good computing practice the recipient should ensure they are actually virus free. - 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] -- There are many ways of going forward, but there is only one way of standing still. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Clustering in Tomcat
I don't know the answer to that, but here's a Cluster/Session Replication HOWTO, in case it might help. http://jakarta.apache.org/tomcat/tomcat-5.0-doc/cluster-howto.html Regards, Faisal On Thu, 25 Nov 2004 09:36:52 +0530, Gaurav Vaish [EMAIL PROTECTED] wrote: Hi, I am facing an issue with Tomcat Clustering. With the details provided on the website, I implemented it. The session (session-ID) is not lost when the machine serving it goes down however all session-data (attributes in session) is lost. What can be the issues? To start with, I had two Tomcat on the same machine with only their port numbers changes (4001 and 4002 for cluster). -- Happy Hacking, Gaurav Vaish http://gallery.mastergaurav.org --- - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] -- There are many ways of going forward, but there is only one way of standing still. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Performance issue
So what's your complaint? That it's GC'ing too much? I don't know if its GCing too much or not. I haven't any idea what the ideal GC should be like. But thanks for your comments on that. That it's not handling enough concurrent clients? Those connection reset messages that were in your log? Do you have defined performance goals are you simply trying to make it not sluggish ? Those connection resets - I don' t know if they're normal or not. Should I be worried about it? I guess I'm trying to find the *reason* why this performance issue exists. Once I'm able to identify it, hopefully i'll be able to do something about it. Regards, Faisal - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Performance issue
Hi, thanks for the insights. I've done what you suggested, and the load does look better, but still sluggish. The machine's are not swapping, simply JVM's max memory is 2GB. Our RAM is 4GB. Any attempt to assign larger max memory to java, will cause failure in starting tomcat. A friend has tried doing this on FreeBSD and Gentoo Linux as well with the same results. By the way, we are having connection timeouts at high peaks: [EMAIL PROTECTED] root]# tail -F -n 100 /usr/local/tomcat/bin/wrapper.log | grep connection jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:33:37 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:33:41 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:33:44 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:33:45 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:33:52 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:33:55 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:34:11 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:34:17 | INFO: connection timeout reached jvm 1| INFO: connection timeout reached INFO | jvm 1| 2004/10/11 15:34:23 | INFO: connection timeout reached On another server, were getting these messages: INFO | jvm 1| 2004/10/12 10:33:11 | INFO: server has been restarted or reset this connection INFO | jvm 1| 2004/10/12 10:33:13 | INFO: server has been restarted or reset this connection INFO | jvm 1| 2004/10/12 10:33:13 | INFO: server has been restarted or reset this connection INFO | jvm 1| 2004/10/12 10:33:13 | INFO: server has been restarted or reset this connection INFO | jvm 1| 2004/10/12 10:33:59 | INFO: server has been restarted or reset this connection And here's a GC log: 43421.932: [GC 1704578K-1569670K(2073856K), 0.8122640 secs] 43436.720: [GC 1756038K-1612010K(2073856K), 0.7409810 secs] 43449.488: [GC 1798378K-1648439K(2073856K), 0.6549710 secs] 43453.472: [GC 1834807K-1679482K(2073856K), 0.5872020 secs] 43481.034: [Full GC 1865850K-809682K(2073856K), 8.6202320 secs] 43502.707: [GC 996050K-851676K(2073856K), 0.5027510 secs] 43523.369: [GC 1038044K-879605K(2073856K), 0.5389870 secs] 43542.764: [GC 1065973K-883973K(2073856K), 0.2121380 secs] 43569.000: [GC 1070340K-895005K(2073856K), 0.1536880 secs] 43591.007: [GC 1081336K-917841K(2073856K), 0.4391310 secs] 43619.284: [GC 1104209K-953106K(2073856K), 0.6477030 secs] 43658.215: [GC 1139474K-985160K(2073856K), 0.5575750 secs] Any ideas? Regards, Faisal - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Performance issue
Hi, I have 4 *really* sluggish Tomcats. All 4 are load balanced: Tomcat Version:4.1.30 Java: Sun 1.4.2_02 SMP: 4 CPU Xeon 2GHz Memory: 4GB OS: Redhat AS 3.0 Users can reach 500 at peak time (about 150 per server). It takes about 2-4 minutes to load. And we're also having session time-outs. Previously, JVM died quite often, so now we're using a wrapper to restart it after it crashes. I have 128M initial memory, and 2GB max memory for JVM. Is this normal? Is this a Tomcat issue, a tuning issue, or a very badly written app? Appreciate the help. Regards, Faisal -- There are many ways of going forward, but there is only one way of standing still. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Performance issue
The OS responds fine even when tomcat responses are slow. The kernel is configured (max connections, file descriptors, semaphores, shared memory, etc). Though i'm not sure whether the tuning is optimal or not. I think its probably its how the app manages connection, and probably GC? DBCP and Cayote HTTP1/1 isn't configured by the way. Would these help? If I configure DBCP, would it effect the code? Sorry for the lame questions, I'm considerably new at this. And thanks for all your fast responses. Regards, Faisal On Wed, 6 Oct 2004 14:52:13 +0200, Nitschke Michael [EMAIL PROTECTED] wrote: It could be that you catch a maximum open handles error. If you are using the standard linux distribution it is set to 1200 connections. And if you have 500 users and every user opens an connection you are up to 1000 and the remaining 200 are very quickly used up. The result is that all java app's (tomcat, console, what ever) crashes instantly. But the other system tasks keep running. You should recompile the linux kernel with about 12,000 connection or more (max is 65000, we use this). You could also use the latest kernel which has no limit for connections. mfg Michael Nitschke -Original Message- From: Filip Hanik (lists) [mailto:[EMAIL PROTECTED] Sent: Wednesday, October 06, 2004 1:49 PM To: Tomcat Users List Subject: RE: Performance issue sounds like an App problem, cause your hardware is sure powerful, assuming you configured your connector threads to handle enough concurrency Filip -Original Message- From: Alex [mailto:[EMAIL PROTECTED] Sent: Wednesday, October 06, 2004 6:20 AM To: Tomcat Users List Subject: Re: Performance issue How big is your average session? On Wed, 6 Oct 2004, Faisal Abdullah wrote: Date: Wed, 6 Oct 2004 17:20:05 +0800 From: Faisal Abdullah [EMAIL PROTECTED] Reply-To: Tomcat Users List [EMAIL PROTECTED], Faisal Abdullah [EMAIL PROTECTED] To: Tomcat Users List [EMAIL PROTECTED] Subject: Performance issue Hi, I have 4 *really* sluggish Tomcats. All 4 are load balanced: Tomcat Version:4.1.30 Java: Sun 1.4.2_02 SMP: 4 CPU Xeon 2GHz Memory: 4GB OS: Redhat AS 3.0 Users can reach 500 at peak time (about 150 per server). It takes about 2-4 minutes to load. And we're also having session time-outs. Previously, JVM died quite often, so now we're using a wrapper to restart it after it crashes. I have 128M initial memory, and 2GB max memory for JVM. Is this normal? Is this a Tomcat issue, a tuning issue, or a very badly written app? Appreciate the help. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --- Incoming mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.768 / Virus Database: 515 - Release Date: 9/22/2004 --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.768 / Virus Database: 515 - Release Date: 9/22/2004 - 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] -- There are many ways of going forward, but there is only one way of standing still. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Cluster Pure Tomcat with Hardware Load Balancer
Here are 2 interesting articles: http://www.onjava.com/pub/a/onjava/2004/03/31/clustering.html http://www.onjava.com/pub/a/onjava/2004/04/14/clustering.html?page=1 Since you do have a hardware load balancer, the article may not make much sense. However, if your concern is mainly session persistency, I guess you should refer to your hardware load balancer vendor for possible configurations to support megaproxy ISPs. Unless you want to opt for Tomcat's clustering feature, it would replication the session throughout all cluster nodes, which *could* solve your problem. Regards, Faisal Abdullah On Mon, 4 Oct 2004 18:09:19 -0400, MITCHELL TEIXEIRA [EMAIL PROTECTED] wrote: Hello to the group - I am desperately searching for information about clustering Tomcat 5 in a pure environment without Apache in front of it. I have a hardware load balancer in front of two Tomcat servers that I want to join in a cluster in order to avoid/reduce broken shopping cart sessions experienced by customers coming from so-called megaproxy ISPs like AOL. Since we can't depend on SSL session ID via the hardware load balancer (thanks to Microsoft IE5+) we believe that we must make clustering work. The problem is scarce/non-existent web documentation of Tomcat Clustering. Any help is appreciated. MitchellT - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] -- There are many ways of going forward, but there is only one way of standing still. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: Drops and Spikes in Tomcat responses
On Tue, 05 Oct 2004 14:03:48 -0700, Kaleb Pederson [EMAIL PROTECTED] wrote: We are using a commerical servlet that seems to keep getting overloaded. We wrote a little script that monitors a number of things about the process - the number of open file descriptors, sockets, java threads, mysql connections, misc. hit types/second and memory sizes. Is it possible you share that script you wrote? I'm having performance problems too, and I'm really new with all this. I need all the help I can get. Thanks. - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
