Hi,

I am having a strange problem. Some of my linux systems (running Redhat linux) are hanging with no error messages on screen or in /var/log/message or dmesg file. I am not sure what is happening with these systems. Can anyone suggest what could be the problem or hoe to debug it?

thanks

Usually denotes some type of hardware issue. The only real way to figure it out would be to log kernel messages ( even then, that may not work )...Personally, when this continued to happen to me, I figured it for a video issue ( after several weeks of testing )...but that's me.

It's also possible the systems you have RH installed on exhibit certain abnormalities...you may want to search to see if the equip you're using for linux is supported.

What kernel version are you running? What type of server? I have similary problems with Redhat 8 stock kernel 2.4.18 on a Dell PE2600. The problems when away with compile kernel-2.4.20. I am currently back to a RedHat kernel 2.4.20-24.8. My problem was tied to the disk.
When you reboot check sar to see if there was any unussual loading just before the system hung (CPU load, disk activity memory).
Let me know what happens

I am running Redhat 7.1.3 on 2.4.9. These are IBM Thinkcentre-M20 machines (P4)

Why are you running such an old version of Redhat. I would at least upgrade the kernel if you can. Did you just set these machines up, or have they been running for a long time and are just now experiencing problems?

Its not possible to upgrade these systems as they are running specific software. These systems are running for quite some time and recently I have started seeing these problems.

These systems are running for quite some time and recently I have started seeing these problems.
What do the upgrade logs/logrotated system logs say?
Can they get into runlevel one or not even that?
Do you keep an admin log for these systems?
Where there any other anomalies?
Did you run HW tests?


I think you should be much and much more verbose on system specs and what you tried to resolve issues etc etc. With all due respect, but I get the idea it's dragging on a bit because of lack of detailed info.

You really should upgrade the kernel.
There are severaly security valuablilities in linux kernels prior to 2.4.23.

You haven't changed anything lately? Added new software or hardware etc...
Is nore than one of the machines experiencing this problem out of the blue?
Are the machines on a secure network?
Do end users have accounts on the box or is it used as like a http server,database server,etc..?

If many machines are suddenly experiencing this problem, andyou haven't added any hardware to the machine. I would say it is not hardware related.

If and you haven't added any software or changed configurations around on these servers maybe you machines have be compramised.

Thanks for all your inputs. I have already run hardware diagonistics on these systems, Their are no Bios error messages etc. These systems are on network and are used to run specific type of jobs. The users of these have changed recently, that could have been causing issues, but I was not sure how to debug the issue.

Does this mean you fixed it? If so, what was the cause (details plz)?

Nope I haven't fixed it yet. Its very random. It works fine for few days and suddenly the system is down with no errors. I am trying to focus on what type of processes ae running when the system goes down. The hardware diagonistics doesn't report any problem. Load testing of systems doesn't produce any errors.

With all due respect, but again this thread is dragging on because of lack of detailed info.
If you want us to be able to help you, it's crucial you share info with us.