Topic: Can switching the power off directly on the MB via power lead damage (Read 1263 times)

Thanks! Did not know 'bout that one.

That is a small but realistic concern.  Newer drives do much better.  They have ability to auto-park using small capacitor to drive the actuator to the park posistion and the drives inertia to park before drive stops spinning.

NOTHING, you'll be incredibly unlucky lose data even, well, maybe 0.001% with windows, 0.000000000000000000001% all other O/S's hehe

Seriously though, Nothing wrong with doing that at all in my 15+ years of experience of 'hard resets'

Not 100% sure but it might not be good for the HDD, when during sudden power loss the heads don't return at their places and land on the platters instead, scratching them.

It shouldn't lead to any hardware damage, as long as you don't do it hundreds times a day.

Thanks for your reply. 

After rooting around I've got more detail from the xorg log file. /var/log/xorg.0.log (for those of you who may come across this searching).

The error message says "Invalid ATI BIOS from int10, the adapter is not VGA-enable. That error comes from the fglrx driver. I'm not sure if I mentioned this in the first post but just to be sure it's an MSI Radeon 6990x3 that I have in my box.

Thanks again.

No.  A cold shutdown can corrupt file systems due to files being in use but it wouldn't have any effect on hardware.

I have two machines which both seem to be having the same issue.

No OpenCL devices found when I try to run my miner. I'm on Linuxcoin 0.2b final and I have been mining with these machines up until recently when I started changing things around software wise - many times I would kill the power by holding down the power button or pulling the lead.

I'm booting from a USB pen drive and have tried a fresh linuxcoin image on the pen drive.

It has been saying "vgaarb this device is not a VGA device on bootup." Is vgaarb not for management of legacy devices only?

My xorg.conf file is the same as another identical rig to it which is running perfectly. Also, x server is running fine on this problem machine.

Maybe vgaarb is not the problem as it's not coming up in the boot messages anymore...strangely. I suppose I'm going to have to root around to see how openCL works and how to troubleshoot.

I've tried reseating the RAM, clearing the capacitors and reseating my 3x6990 graphics cards. It's using an MSI 890 FXA GD70 board with a Corsair 1200AW PSU. When I run a memtest I get error 0604 (I think it is). I've swapped the memory into different slots. It looks like the memory...but I'll have to go and buy online, wait for post etc., after that it may turn out that it's the MB....etc. However it does boot into Linuxcoin ok, everything seems to work fine. There is another issue where it does not initialize eth0 but eth2. Before that it was only showing localhost and eth3 - dhcp wasn't working. The reseat of the memory appears to have resolved that abit.

Hmmmm.