RE: [suse-amd64] amd production server crashes under a network lo ad. how to debug?
What version of Tomcat What version JVM What version of Apache (I am assuming version 9.1 of SuSE) Are you running Tomcat with the "-server" option with addition RAM allocated to the JVM (> 256MB). -----Original Message----- From: slava zimine [mailto:sadko77ch@mail.ru] Sent: Monday, July 12, 2004 10:13 AM To: suse-amd64@suse.com Subject: [suse-amd64] amd production server crashes under a network load. how to debug? Hi 2 al. Please, forgive me. My question is not about suse on amd64 arch because my problematic server in question is 2 x amd athlon M 1.3 Ghz. Yet I find this mailing list highly informative and relevant to server linux and I did not want to register on on suse-linux-e@ list to post one message. My problem is that since a month i experience an increased number of network inaccessibility of my production 2 x amd athlon M server, which serves a java servlet based e-commerce webapp on tomcat+apache with a small load (200 visitors per day). It is a remote server in another country, in which I can only ssh. the problem: under increased network load (like try to download from the web suse updates) the system crashes, and becomes network inaccessible. When i ask an ISP operator to go to see my server, he cannot log in into it, the monitor stays black. Only the system reboot helps. (the system always rebooted ok so far). I see nothing in my logs /var/log/(local)messages. The OS: suse linux 8.0 professional (runned ok, for 2 years). () no online updates were performed on the server. I already changed a network card (3Com) but the crashes persist. Question: how to debug & narrow the problem? How to make sure it's not a disk subsystem? (all server partitions(ext3) are on hardware raid5 on adaptec2100S. or a motherboard (tyan) ? How to see which part of the kernel crashes? I would like to understand if this is hardware or software problem, before flying there and installing a suse linux enterprise linux. Thank you for your comments. kind regards. -- Check the List-Unsubscribe header to unsubscribe For additional commands, email: suse-amd64-help@suse.com
I don't think the crashes are related to jvm & tomcat. the higher usage of tomcat webapp did produced crashes. but so did, the download by wget of 2Gig directory with suse online updates, as well as a run of a webalizer on apache yearly logs with dns resolution (= lots of DNS requests) tomcat 4.1.24, jvm j2sdk1.4.1_05 CATALINA_OPTS="-Xmx900m -Xms800m" kind regards slava On Mon, 2004-07-12 at 17:22, Kevin.Wilson@comtrol.com wrote:
What version of Tomcat What version JVM What version of Apache (I am assuming version 9.1 of SuSE)
Are you running Tomcat with the "-server" option with addition RAM allocated to the JVM (> 256MB).
-----Original Message----- From: slava zimine [mailto:sadko77ch@mail.ru] Sent: Monday, July 12, 2004 10:13 AM To: suse-amd64@suse.com Subject: [suse-amd64] amd production server crashes under a network load. how to debug?
Hi 2 al. Please, forgive me. My question is not about suse on amd64 arch because my problematic server in question is
2 x amd athlon M 1.3 Ghz.
Yet I find this mailing list highly informative and relevant to server linux and I did not want to register on on suse-linux-e@ list to post one message.
My problem is that since a month i experience an increased number of network inaccessibility of my production 2 x amd athlon M server, which serves a java servlet based e-commerce webapp on tomcat+apache with a small load (200 visitors per day). It is a remote server in another country, in which I can only ssh.
the problem: under increased network load (like try to download from the web suse updates) the system crashes, and becomes network inaccessible.
When i ask an ISP operator to go to see my server, he cannot log in into it, the monitor stays black.
Only the system reboot helps. (the system always rebooted ok so far).
I see nothing in my logs /var/log/(local)messages.
The OS: suse linux 8.0 professional (runned ok, for 2 years). () no online updates were performed on the server.
I already changed a network card (3Com) but the crashes persist.
Question: how to debug & narrow the problem?
How to make sure it's not a disk subsystem? (all server partitions(ext3) are on hardware raid5 on adaptec2100S.
or a motherboard (tyan) ?
How to see which part of the kernel crashes?
I would like to understand if this is hardware or software problem, before flying there and installing a suse linux enterprise linux.
Thank you for your comments.
kind regards.
participants (2)
-
Kevin.Wilson＠comtrol.com -
slava zimine