Since moving to 11.1 from 11.0 I have been getting random reboots. I’m really not a hardware guru and would welcome any advice / starting point.
I’m suspicious of Flash player being involved as it seems to happen when playing bbc.co.uk/iplayer. I had switched from FF to eliminate that and have had it happen now with Opera too. I have done complete re-install and re-format of drives and partitions.
My only other concern is that it may be one of the HD’s (in particular an older Maxtor). How would I test this?
The way it happens: Is just as if the power went off. Everything just drops. And immediately it kicks in to boot up again.
TBH, although I rarely say this, my first thought is that it sounds like an issue with the hardware itself.
If the kernel crashes, it will generally hard crash and freeze the system by design. I’ve had lockups before, and I’ve had my system automatically power down from overheating, but I’ve never seen a reboot.
Failing hard drives will cause issues, but they generally cause system instability, they shouldn’t cause a reboot. You can check the status of your hard drive with smartmontools, you’ll probably have to install it via Yast. SMART monitoring in hard drives sometimes throws false-positives, but it rarely fails to report issues when the drive is failing.
If you happen to dual-boot that system, have you experienced any issues with other OSes?
You could also take a look at /var/log/messages the next time your system reboots, to see if the log captured any error messages prior to the crash.
You’re going to need a little more info to figure out what is crashing it, and it’s usually easiest to rule out potential causes. For instance, you can try stress-testing your system to see if you can force a crash (such as running glxgears fullscreen for a while, or something CPU intensive like compiling a kernel) to see if your system is failing under load, which could be the case. An overheating CPU or GPU can cause unpredictable behavior, though the BIOS should shut down the system when they reach a certain threshold.
Thanks for your response there Kev. I am currently in ‘Process of elimination’. Seeing that all occurrences to date seem to have been when running Flash Player - I have just rolled back Flash to the last version that had caused no issue. Though that was using OS 11.0 _64. Which in fact of course was a bit more stressful on the system using the plugin for 64bit flash. It could get the CPU fan stepping up a gear. But never any sign of overheating there or with the GPU either. CPU seems to run a constant 40-42 degrees Celsius.
I just opened my GPU monitor and it was @ 42 degrees. The nVidia settings show a greyed out ‘shutdown threshold’ of 125 degrees. The temp quickly lifted to 50-52 with glxgears at fullscreen. My GPU has been modified. The original and very noisy fan has been replaced by a large heatsink (Zalman VNF100). But never any issue with this, and it is cleaned regularly.
I mentioned the ‘old’ HD. Nothing has really ever given me reason to think there is a problem here. Apart from the fact that it is a Maxtor and can sound rather like a ‘bag of nails’ - it has been fine. Previous file checks have revealled some inconsistancy in the file system, but the drive had not been formatted for years, it held swap and /home and grub was on the MBR. So, most recently I have formatted the drive, it now has swap and a large ext3 storage partition used for VM’s and grub is still on MBR.
I do have XP on another newer drive along with my openSUSE partitions. And it would be my intention to try it. It may help eliminate the flash question, as a reboot in XP would tend to suggest hardware.
Glxgear is still running, GPU is @55 now and the CPU fan is picking up a little, though sysinfo still shows a temp of 40 there.
It’s the rebooting thing that gets me. If it was hardlocking or something, I could see that.
Is it possible that there are power fluctuations? Maybe a loose connection in the power cord, or the PS connectors in the system?
After that, I’m dry. Narrowing it down to software would be the next step, and you’re doing that, but I’d still be surprised if a user-app could cause that.
The fact that it has started happening immediately after installing OS 11.1 is why I am less inclined to suspect hardware. I have not been inside the box for about six months, so there is unlikely to any duff connections. I know it can happen, but this box is a fairly recent self build and all the connections are fresh. I’ll keep it in mind though should I take the side panel off and make sure I freshen the connections.
I’ll post back here as I progress, but it may be a long winded affair. As re-boots are not frequent and seem to be dependent on me using flash, which I don’t do all the time.
Hi
Do you have an intel cpu? If so it may be the iTCO_wdt (watchdog) and
iTCO_vendor_support modules that seem to have appeared in this new
kernel… I would rmmod them and see how it goes.
–
Cheers Malcolm °¿° (Linux Counter #276890)
openSUSE 11.1 x86 Kernel 2.6.27.7-9-default
up 10:39, 1 user, load average: 0.22, 0.26, 0.27
GPU GeForce 6600 TE/6200 TE - Driver Version: 180.22
Well like someone already posted in a not so very informative post earlier today in another thread, power supplies can cause all kinds of strange issues.
It could be shutting down due to flash putting the CPU to work, do you run anything else that’s CPU intensive without problems?
When I use to Overclock my systems a lot , I found that crashes as you described are Memory related , a bad memory stick , when it is written to at a specific block it crashes , run memtest bet you find a bad block !
It could be shutting down due to flash putting the CPU to work, do you run anything else that’s CPU intensive without problems?[/quote]
Yes I do. Often run video encoders - and VBox can get the rev’s up too.
But I may be wrong about flash. I had switched back to an old version 9, which I had never had issue with in OS11.0 but did get the reboot today during a bbc flash program. It may just be coincidental that it happens during a long flash browser session. I had glxgears running full screen as suggested by Kev earlier, with no reboot.
I’m tempted to loose the old Maxtor which is used as swap, VM store and bootloader.
If I grab a bit of space on my newer WD HD for swap, install grub to the WD and reconfigure the fstab for boot. It will be easy enough.
>
> Thanks for these suggestions folks. It’s AMD Malcolm.
>
> I try the other suggestion re:Mem.
>
> Crash has happened with the old flash too, so may rule that out.
>
>
and rum the different diagnostics from the bootable CD, there are tools there for loads of different stuff have been using it for years and it has found plenty of hardware probs for me
Ok, can you check your RAM specs for voltage and can your BIOS tweak
the voltage? I found on my SLED system (X2 4400) I had to up the voltage
from memory (no pun intended) .2 or .3V to get it running properly.
I would look at running prime on it as a test to check for memory
errors and see what happens, memtest is ok, but you really need a
load on it.
Your running 32bit, I also wonder if it’s worth installing the kernel
source and looking at the .config file and see if there is anything
there is anything that may cause a problem?
–
Cheers Malcolm °¿° (Linux Counter #276890)
openSUSE 11.1 x86 Kernel 2.6.27.7-9-default
up 1 day 5:33, 4 users, load average: 0.15, 0.15, 0.24
GPU GeForce 6600 TE/6200 TE - Driver Version: 180.22
