Server LOCKUP, if You have XEON server please HELP !!!

[VIPsNET]

Hey guys,
I need some help from experienced sysadmins. I have 6 DELL Dual Xeon servers from 2Ghz to 2.4Ghz, (and one 3.2Ghz but I just have it for 2 weeks), SCSI HDDs and RHEL, anyways from time to time all those servers lockup/freeze for no reasons at all, I've seen TOP command output from the second it locked up, logs shows no errors etc. I Googled and read on other forums that some users experience this issue as well, it was well known a year ago when there was a bug in SCSI card driver but it has already been fixed. What I would like to know if You guys who have DELL Dual Xeon servers experience server lockups as well ? The lockups are totaly random ie. it can heppen twice a day then work for 30 days... Here is my configuration:

Dell Dual Xeon 2Ghz - 2.4Ghz
RAM 2GB
2GB SCSI HDDs
RedHat Enterprise 3 with custom Kernel 2.6.11 (the default Redhat Kernels causes more frequent freezes).
cPanel as control panel

all servers are located in EV1.

The load average on the servers vary from 1-2 on some of them to 6. Sometimes server lockup with load 1-2 and sometimes with 8.

I suspect that this has something to do with hard drives, SCSI card, Kernel.

I used to own P4 and Celeron servers with IDE drives and they were very stable, those Xeons lockups too often and it is high time to do something.

My friend has 5 Xeon servers and they lockup too from time to time.

Can You guys who have Xeon servers tell me:

1. If You experience the lockups ?
2. If so, how often ?
3. What hardware/software do You have
4. DC your server is located

I think of changing SCSI drives to IDE drives but I'm not 100% sure if this would help.

I would love to solve the lockups, I can stand lockups every 2-3 months but not every 1-3 weeks.

[jackal]

2.4.21-20.0.1.ELsmp No issues plus remove auditd if it is loaded

[VIPsNET]

Do you have SCSI drives and if You know let me know name of your SCSI controller.

[Blue|Fusion]

2.4.21-32.0.1.ELsmp working fine for me. At The Planet, (Total Control P2800SS) DLLSTX4. Dual Xeon 2.8GHz, RHEL3, no RAID, three 73GB 10,000RPM SCSI. So far, no problems with lockups. A friend, running a similar machine at the same place also reported no problems.

I did hear that the lockups were almost always with RHEL3 servers, but I never had issues with it. Perhaps you can test something out with RHEL4?

[chirpy]

As jackal has said, do make sure that you do not have the laus rpm installed for auditd and if you have remove it and perform the cron fixes that are mentioned in other threads for it on these forums.

[VIPsNET]

I do not have auditd and laus, I'll try 2.4.21-32.0.1.ELsmp kernel. What is your load average guys? my servers have 2 to 6-8 at peak times.