1

I have a Gateway GM5474 which is running XBMCbuntu on a 500GB Hard drive (the originial I believe or maybe slightly newer hard drive). Little background: The system is still stock, and I was having issues getting any linux kernel to install and I considered it a miracle that it installed at all. I tried all three hard drives I have until it installed on the 500GB. (the others are 500GB and 1TB)

It was working fine but now it randomly freezes up, on reboot there is nothing logged to syslog, nor is there anything in dmesg indicating that there is an issue.

I thought it may have been due to the screensaver as I could not ping it after the screensaver had been initiated for longer than 20 minutes. I also can't turn off the screensaver for whatever reason, I tried the command line method and the gconf method and neither affect the screensaver. [EDIT:] It was not always the screensaver causing the issue, it seems to stall after a solid hour has elapsed.

I would like some advice but if the general consensus is to reinstall using something suited to older hardware then so be it.

Cole Busby
  • 197
  • Is the screensaver still running? Assuming you have something animated I mean. – terdon Sep 19 '13 at 15:45
  • I couldn't change it from the fade-to-black screensaver...tried that too. – Cole Busby Sep 19 '13 at 15:50
  • Is it still on? I mean, can you hear the fans, is the HDD light blinking etc? Do you have it set to suspend or hibernate after 20 minutes of inactivity perhaps? – terdon Sep 19 '13 at 15:52
  • The fans were still running and but the HDDs showed no activity but were still spinning. It was as if the hibernate function initiated but never completed. Little side-note, it freezes without the screensaver as well. – Cole Busby Sep 19 '13 at 15:53

1 Answers1

1

First, try to run memtest86 to establish whether CPU/mainboard/RAM/power supply are working properly. If you get errors, it could be any one of those - I'd say a 80% likelihood it's a RAM issue, 19% likelihood it's a mainboard issue, and a 1% likelihood that it's power supply issue. It's most probably not a CPU issue. If you get errors and you have multiple memory modules, try removing all but one and test again.

I tried all three hard drives I have until it installed on the 500GB. (the others are 500GB and 1TB)

That sounds rather funky. Luckily, hard disks are the one thing in a computer that you can test very reliably with SMART selftests. Start a selftest as root on a disk with

smartctl -t long /dev/sdX

where /dev/sdX is the hard disk you want to test. You can do this while the disk is in use; the selftest runs in the background, won't block the disk and won't damage any data. It will take a while though, depending on the hard disk size. smartctl will tell you exactly how long when you start it. Don't reboot or shutdown your computer as this will interrupt the selftest so that you would have to start it again from scratch. When it's done, query the result with

smartctl -a /dev/sdX | less

Scroll down to the SMART Self-test log block. If the topmost test says Completed without error the disk is fine, if it reports any error, throw the disk away. The results have been 100% accurate for me, there's no uncertainty involved like with other computer components.

Martin von Wittich
  • 14,187
  • Ran Memtest86+ and unfortunately I should've started this first, I switched my configuration to individual reports and it started producing errors, I took out one stick of RAM and voila no more errors are being produced, I'm trying another slot to diagnose a bigger motherboard issue, if that pans out then I'll be happy otherwise It's off to amazon. Do you know of where to get the cheapest RAM. – Cole Busby Sep 20 '13 at 03:10
  • Thank you for your help! As previously stated I attempted to set the suspected bad RAM in a different slot, memtest did not have errors any longer, I believe this is because the slot has gone bad but only time will tell. – Cole Busby Sep 20 '13 at 15:15