Hi!

 

I have got a 5 disk UNRAID system running at the time, the system seems to be stable (more than 8 hours memtest with no errors) and I am getting unexpected crashes (the system resets and start rebooting). This is not happening all the time but it only happens when disks sleep and I am trying to access one of them or the whole system.

I am sharing a syslog but unfortunately its after the crash...

 

I have got hdd backplanes installed a Gigabyte G33M-DS2R, 2Gigs of OCZ, and a Corsair VX550 single rail psu if this helps...

 

Any ideas?

syslog.txt

I have not looked at the syslog, but this kind of thing almost always points to a hardware problem.  My first suspect would be the PSU and then maybe the motherboard. The VX550 you have is a good PSU but there can still be problems with it.  My next guess would be the backplane, but it would be the first thing I would check.

 

Take the drives out of the backplane and hook them up separately.  Then start using the machine like normal, if it crashes start by removing the PSU next.  Do a process of elimination to figure out what might be causing the problem.

I ll check it with a new psu tomorrow. The strange thing is that is seems to be stable by passing 8+ hours of memtest...

I ll check it with a new psu tomorrow. The strange thing is that is seems to be stable by passing 8+ hours of memtest...

But when running memtest, you are not spinning up the drives.  That takes a lot more power.

 

Joe L.

You are using OCZ memory - many of these are marked and sold as "overclockable" and usually require a higher voltage that the standard 1.8V.

 

Check your BIOS setting , flash to the latest release and where you are there read and disable the HPA "feature" famous on any Gigabyte board.

After that check out your memory specks, set the memory with a manual settings and up the memory voltage to the required level, run memtest again and check to see if everything is OK now.

Thank you all for your good ideas!

I did checked the +12V on the psu and when hard disks spin up it drops from 12.10 to 12.06 so I dont thing that this is the problem but after all I ll try a new psu tomorrow.

About memory timings now. They are all set to automatic in bios but ocz states that they are 1.8V:

http://www.ocztechnology.com/products/memory/ocz_ddr2_pc2_6400_gold_gx_xtc_dual_channel

Also a couple of days before I was using Corsair Dimms with the same results (automatic settings also):

http://www.corsair.com/_datasheets/TWIN2X2048-6400C5DHX.pdf

 

I ll do those checks too and hope I ll find the problem...

Thank you all for your good ideas!

I did checked the +12V on the psu and when hard disks spin up it drops from 12.10 to 12.06 so I dont thing that this is the problem but after all I ll try a new psu tomorrow.

About memory timings now. They are all set to automatic in bios but ocz states that they are 1.8V:

http://www.ocztechnology.com/products/memory/ocz_ddr2_pc2_6400_gold_gx_xtc_dual_channel

Also a couple of days before I was using Corsair Dimms with the same results (automatic settings also):

http://www.corsair.com/_datasheets/TWIN2X2048-6400C5DHX.pdf

 

I ll do those checks too and hope I ll find the problem...

 

Your original system log does not show a crash.  To capture the system log, you need to use telnet to log into the unRAID shell, and then type this command:

 

tail -f /var/log/syslog

 

This will cause syslog entries to get displayed in the telnet window as they occur.  After the crash, you can select/copy/paste the telnet window text & post here.

yeap I know that about syslog. Unfortunately system resets instantly so I cannot capture syslog the way you mention. Still testing though... This time I have increased Vdim to 1.9V. bcbgboy13 was right (also waiting for a new psu), many people state that they need to push the Volts a little bit in order to get stable, even though dimms suppose to work with 1.8V.