Kernel Errors Present

singeX · 08-07-2007, 09:59 PM

After every server boot up I get the following error from logwatch:

Code:

WARNING:  Kernel Errors Present
    Contact your BIOS vendor to see if the E752x error registers can be ...:  1 Time(s)
    e752x_edac: probe of 0000:00:00.0 failed with error -12 ...:  1 Time(s)

The server is running but some services hang or run very slow every now and then. A Google search didn't help much. All I know is that it has something to do with the RAM which would probably explain why some things are running painfully slow. I just did a yum update to try and solve the problem. The downloads went fast but the updates took a few hours to do. "top" doesn't show anything significant, averages 0.5 - 1.0% cpu load on each. Also, when I use the "reboot" command it takes an hour or so before it actually reboots. I assume all of this is related but I'd like to hear other thoughts. Oh, and the server ran perfectly on Fedora Core 4. I'm hoping it's a kernel bug and that booting off the 2.6.18-8.1.8.el5 kernel will solve it. Either way, I'll post an update here when it finally reboots.

OS: CentOS 5 x86_64
Kernel: 2.6.18-8.el5
CPU: Dual Xeon (single-cores) 3GHz
RAM: 4GB PC2700
HDD: RAID 5 3ware 9550sx

singeX · 08-08-2007, 07:10 PM

Booting from the 2.6.18-8.1.8.el5 kernel did not solve the problem. Instead, I gained two more kernel errors:

Code:

hdd: packet command error: error=0x54 { Ab ...:  2 Time(s)
hdd: packet command error: status=0x51 { D ...:  2 Time(s)

singeX · 08-27-2007, 04:24 PM

I don't know for sure if this was related to the Kernel errors or not but I added "noapic" and "nolapic" to the kernel parameters. I don't have any noticable problems now.

ahack · 01-31-2008, 11:04 AM

Check for BIOS updates and monitor your system to see if the slowness is from the drives or CPU or what.