Serious help needed, with a poweredge 1950 running debian... random reboots

Question

Ok, so I'm now officially at my wits end with a machine I have.  I've talked with Dell support, and have run all the diagnostics that I can run, but the problem still persists...  Here's the deal.   The Problem:     My PowerEdge 1950 randomly reboots itself.  I can sometime predict that it will happen in the morning, but it's never at the same time of day.  When it reboots, it's doing a HARD reboot, there are no logs of the system shutting down, one second it's up, the next it's down again.  I've even added remote logging to the system, just in case something was getting logged, but not written to disk locally... still nothing.      I've gone through everything I can think of to figure out what the problem is, but still can't see anything wrong with the system.  Since this started happening out of the blue, (the machine was running without incident for about two years), and we didn't make any software changes that I can see.  (I'm the only administrator, so I'd be the only one making changes), I don't suspect this to be a software issue.   I've been monitoring the fans, ambient temp, etc for the past month and they all check out.  The Fans run consistently between 5k -7.5k RPM depending on the fan, and the ambient temp is about 20c, with max spikes to 22c.     To show you what i mean by just restarting.. here's a snipet of the syslog of a reboot the other day...  the times are all messed up, cause I don't think the system CMOS Clock was set correctly... so during boot it goes back a few hours, then NTP fixes the system time at the end of the boot cycle...   Jul 21 09:00:44 servername snmpd[3351]: Connection from UDP: [xxx.xxx.xxx.xxx]:48828Jul 21 09:00:44 servername snmpd[3351]: Connection from UDP: [xxx.xxx.xxx.xxx]:48828Jul 21 09:00:44 servername snmpd[3351]: Connection from UDP: [xxx.xxx.xxx.xxx]:48828Jul 21 09:00:44 servername snmpd[3351]: Connection from UDP: [xxx.xxx.xxx.xxx]:48828---- TIME SHIFT HERE---- due to CMOS CLOCK I THINK.Jul 21 03:07:09 servername syslog-ng[3615]: syslog-ng starting up; version='2.0.0'Jul 21 03:07:09 servername kernel: Linux version 2.6.18-6-686 (Debian 2.6.18.dfsg.1-18etch6) (dannf@debian.org) (gcc version 4.1.2 20061115 (prerelease) (Debian 4.1.1-21)) #1 SMP Fri Jun 6 22:22:11 UTC 2008Jul 21 03:07:09 servername kernel: BIOS-provided physical RAM map:Jul 21 03:07:09 servername kernel: BIOS-e820: 0000000000000000 - 00000000000a0000 (usable)Jul 21 03:07:09 servername kernel: BIOS-e820: 0000000000100000 - 000000007ffa8000 (usable)Jul 21 03:07:09 servername kernel: BIOS-e820: 000000007ffa8000 - 000000007ffb7c00 (ACPI data)Jul 21 03:07:09 servername kernel: BIOS-e820: 000000007ffb7c00 - 0000000080000000 (reserved)Jul 21 03:07:09 servername kernel: BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved)Jul 21 03:07:09 servername kernel: BIOS-e820: 00000000fe000000 - 0000000100000000 (reserved)Jul 21 03:07:09 servername kernel: 1151MB HIGHMEM available.  ....   As you can see... one second it's up, the next it's rebooting.   We've gone through all the Dell System tests and all hardware checks out.  Suspecting that the problem was RAM, I ran a deep test of memory using memtest86.... and still everything checks out.   The system has redundant power supplies, connected via a Y cable.  It doesn't have a RAID controller, it's just using the onboard SAS Controller.   I'm running Debian Etch, with a 2.6.18 kernel.   Can anyone help me see what could be causing this box to reboot in such an odd manner?  Have I just missed something compeltely obvious?   Thanks

gtdaqua · Answer

check the firmware and driver version of your sas6i/5i or perc5i modules. i had random restarts on my ubuntu but for now its ok. only today i figured how to apply the latest drivers. i dont monitor these posts but try upgrading the firmware and drivers first.

because you use debian, dell does not supply drivers natively. go here to compile from the source:

http://www.dellcommunity.com/supportforums/board/message?board.id=sw_linux&message.id=14001#M14001

more details:

http://ubuntuforums.org/showthread.php?t=719556&page=3

PowerEdge OS Forum

Was this post helpful?