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.

I got a Dell Latitude E6520 with NVIDIA GF119 [Quadro NVS 4200M] video card and a SAMSUNG SSD PM810 2.5" 256GB (firmware AXM06D1Q) hard disk.

Occasionally (used to be once every few months, now more like once or twice a week) it crashes where the laptop is still powered and display still shows my desktop frozen in time (mouse won't move, the scrolling graphs in the system monitor applet are not moving). I can't SSH into the laptop remotely. The only thing I can do is turn is pull the power (press and hold the power button).

Invariably, there's nothing in the logs. Everything was running fine with normal operational messages and then nothing, until normal boot messages when I bring it back up.

I've tried two different distros, that came with different kernels, so don't think it's software related.

How do I even begin troubleshooting something like this? Could it be the hard disk? Or RAM?


UPDATE 2012-07-14

Setup a background job to dump dmesg output to a log every 0.1 seconds to catch any dmesg output that might not be getting logged.

Was working on the laptop till about 2am last night, then was out all day, got back on the laptop at 5pm today to find it locked up. Unfortunately, nothing in dmesg that looks like a likely cause:

kern  :info  : [Fri Jul 13 08:53:13 2012] usb 2-1.2: new full-speed USB device number 4 using ehci_hcd
kern  :err   : [Fri Jul 13 08:53:13 2012] usb 2-1.2: device descriptor read/64, error -32
kern  :err   : [Fri Jul 13 08:53:13 2012] hub 2-1:1.0: unable to enumerate USB device on port 2
kern  :info  : [Fri Jul 13 08:53:14 2012] usb 2-1.2: new high-speed USB device number 5 using ehci_hcd
kern  :info  : [Fri Jul 13 08:53:14 2012] Initializing USB Mass Storage driver...
kern  :info  : [Fri Jul 13 08:53:14 2012] scsi6 : usb-storage 2-1.2:1.0
kern  :info  : [Fri Jul 13 08:53:14 2012] usbcore: registered new interface driver usb-storage
kern  :info  : [Fri Jul 13 08:53:14 2012] USB Mass Storage support registered.
kern  :notice: [Fri Jul 13 08:53:15 2012] scsi 6:0:0:0: Direct-Access     Motorola A955             0001 PQ: 0 ANSI: 2
kern  :notice: [Fri Jul 13 08:53:15 2012] scsi 6:0:0:1: CD-ROM            Motorola A955             0001 PQ: 0 ANSI: 2
kern  :notice: [Fri Jul 13 08:53:15 2012] sd 6:0:0:0: Attached scsi generic sg2 type 0
kern  :warn  : [Fri Jul 13 08:53:15 2012] sr1: scsi3-mmc drive: 0x/0x caddy
kern  :debug : [Fri Jul 13 08:53:15 2012] sr 6:0:0:1: Attached scsi CD-ROM sr1
kern  :notice: [Fri Jul 13 08:53:15 2012] sr 6:0:0:1: Attached scsi generic sg3 type 5
kern  :notice: [Fri Jul 13 08:53:15 2012] sd 6:0:0:0: [sdb] Attached SCSI removable disk
kern  :info  : [Fri Jul 13 11:34:42 2012] usb 2-1.2: USB disconnect, device number 5

:-(

share|improve this question
    
Anything suspicious in dmesg? –  thisfeller Jul 12 '12 at 19:15
    
@thisfeller He can't check dmesg at crash time. And afte reboot, it's always clean. –  ott-- Jul 12 '12 at 19:32
    
I need to read more carefully. –  thisfeller Jul 12 '12 at 20:01
    
Most likely the OS took a wild branch. That could be due to bad RAM, or could be a bad motherboard or CPU. And, of course, it could be due to the disk supplying bad data or the disk adapter garbling it on the way in. –  Daniel R Hicks Jul 15 '12 at 3:13
    
The very last line there is from when you unplugged your DROID 2. The laptop continued working fine after that? –  Michael Hampton Jul 15 '12 at 3:58

2 Answers 2

I suppose I'll take a stab at this one. I situations such as these, I ALWAYS like to test all of the components (RAM, HD, Power, bad caps, etc.) before anything else, just because it saves so much time and effort later on down the road. Basically, this set of steps is a good way of covering your bases so you aren't banging your head into a wall later on because you missed something.

Check the computer for any obvious indicators of failure. Leaky capacitors, broken fans, burnt connectors, super hot components, dying video card, anything that is visible is usually a first good step.

Next, there are a variety of bootable disk utilities that you can use to test the health of various hardware pieces. One simple but effective tool I like to use to check Hard Drive health is Drive Fitness Test. This will tool will essentially go through the drive checking for bad sectors as well as performing a number of other checks in the process.

Another simple thing to test for would be faulty RAM, with something like Memtest86+.

share|improve this answer

You can redirect syslogd's output to another syslogd in the network. See http://linux.die.net/man/5/syslog.conf the example for *.* @finlandia to send everything to a remote syslogd.

share|improve this answer

Your Answer

 
discard

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.