Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Over the weekend one of our Beowulf cluster nodes went offline. It's a Fedora 16 machine. When I try to boot it up, it goes through grub, gets to "initializing ramdisk..." and then reboots every time. It also reboots after the "Starting Windows..." logo animates when booting from a Windows 7 CD. I have run MemTestx86 and it has no errors with repeated passes. It's not currently an overheating problem, but some hardware may have been damaged due to heat stress since until this week we didn't have good ventilation for these cluster nodes on the top shelf.

The machine is one of seven i7 2600K mini-ITX nodes with only the built-in video. It on the ASRock Z68M-ITX/HT motherboard. It has been running overclocked to 4.1Ghz with its DDR3 RAM running at 1833 (within specification for this RAM). While trying to diagnose it I set it back to standard clock speeds but it made no difference. I'm pretty certain it's not the hard drive because it passed fsck.ext4 on all partitions when plugged in externally to another one of the nodes and it reboots when booting from CD too.

Is there a way I can increase the level of detail during the Linux bootup so I can see exactly where it is before it reboots?

share|improve this question
    
It's a hardware fault. Likely motherboard or CPU. The only realistic way to diagnose it is to swap out components until it goes away. –  Harry Johnston Apr 3 '12 at 22:04
    
@Harry: but why would a motherboard or CPU failure happen every time at the same point in the OS boot and not at all during MemTest? –  Matt Chambers Apr 3 '12 at 22:06
    
I've seen similar hardware issues before. MemTest behaves very differently to an operating system, so it shouldn't be too surprising if it doesn't use (or only lightly uses) some of the transistors or other low-level components. If the fault is in one of those components, then MemTest will run even though a real operating system fails. –  Harry Johnston Apr 3 '12 at 22:30
    
But basically it's down to process of elimination - it can't be a software problem if it happens when booting from a known-good CD, as well as from hard disk drive. Could conceivably be firmware, I suppose, but hardware failure is more likely. Since it is always happening at a specific point at the boot cycle, it is likely that it is either (a) when the CPU mode is changed, or (b) when a particular piece of hardware is initialized. –  Harry Johnston Apr 3 '12 at 22:35
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.