November 8, 201015 yr All, I continue to have problems where my unRAID server will lock up when I try to write to it. It writes ~200MB and it will freeze and become completely unavailable to Telnet or unMENU or unRAID menu. I had to do a hard reboot (flipped the power switch on PSU) and when it started back up it sees my Parity drive a New... What do I do from here? Say yes and let it rebuild? I will run a tail once I start so we can see what kind of errors I encounter.
November 8, 201015 yr Author Parity Sync started with a tail running... I will report back in 800 minutes or so
November 15, 201015 yr Author So here is a tail of the system log as I rebooted my server (hard) and it started up a parity check. It finishes parity check and I then try to copy over files to one of the disks. So I see the CPU errors... what do they mean? syslog.txt
November 15, 201015 yr So here is a tail of the system log as I rebooted my server (hard) and it started up a parity check. It finishes parity check and I then try to copy over files to one of the disks. So I see the CPU errors... what do they mean? Do you have a heat-sink on your CPU? Did you use heat-sink compound between the CPU and the heat-sink when you installed it? Did you install an add-on to track CPU temperature? (I've not seen those warning/threshold messages before)
November 15, 201015 yr Author So here is a tail of the system log as I rebooted my server (hard) and it started up a parity check. It finishes parity check and I then try to copy over files to one of the disks. So I see the CPU errors... what do they mean? Do you have a heat-sink on your CPU? Did you use heat-sink compound between the CPU and the heat-sink when you installed it? Did you install an add-on to track CPU temperature? (I've not seen those warning/threshold messages before) Joe, This CPU/MOBO has been in place for 2+ years. Yes, I used compound between the heat-sink and CPU. I have never seen these errors ever. I have not installed a add-on or anything....
November 15, 201015 yr A heat issue is causing the CPU to throttle itself, which may be leading to the freezing issue. I would take the board out and see if any caps are blown, bulging, or just look odd. It would probably be a good idea to reseat the CPU heatsink since that is where the errors are coming from.
November 15, 201015 yr A heat issue is causing the CPU to throttle itself, which may be leading to the freezing issue. I would take the board out and see if any caps are blown, bulging, or just look odd. It would probably be a good idea to reseat the CPU heatsink since that is where the errors are coming from. It could be that one or more of the heatsink attachments has come loose. As prostuff1 said, I'd look there first. If you do replace the heatsink compound remember to use very very little (an amount about the size of a grain of rice, smoothed to cover the entire surface of the CPU with the edge of a business card. Lots of illustrations on the web of how to properly do it. Many use WAY too much heat-sink compound)
November 15, 201015 yr Author Thank you. I went and removed my heat sink and cpu, cleaned them off and applied Arctic Silver 5... Booted back up and parity check started. I will advise later on if I encounter and other problems... THANK YOU
November 16, 201015 yr Author Damn... still the same problem when writing to the drive. This time there were no errors on the tail about the CPU, which is good news, but something is messed. I will have to shut it down (hard) as the array is not accessible when it gets to this state. I wonder what the heck could be wrong....
November 16, 201015 yr Author It might also we worth noting that it writes for about 296MB and that is when it hangs...
November 16, 201015 yr Damn... still the same problem when writing to the drive. This time there were no errors on the tail about the CPU, which is good news, but something is messed. I will have to shut it down (hard) as the array is not accessible when it gets to this state. I wonder what the heck could be wrong.... You could run tail -f /var/log/syslog on the system console to see if you can capture the cause of the failure. Prior to doing that, I'd run a memory test for several cycles to ensure it is not the culprit.
November 16, 201015 yr I have experienced these kinds of events before. It has always been a hardware fault for me. Identifying which piece is at fault is the hard part. Having extra parts is the key to easily finding problems. Replace RAM (or use a single stick) Swap power supplies Swap power splitters Swap SATA cables Bypass drive cages Swap controller card(s) Swap motherboard Rebuild USB stick Put in old disks and try running with the same hardware Put in new parity drive and try building parity with a subset of your disks The order I do things in is determined by how time consuming they are. If you have old/extra disks troubleshooting without having your real data at risk with reduce the stress levels.
November 16, 201015 yr Author Joe: I've run Memtest for 20+ hours and that does not seem to be the problem. Once this Parity check finishes I'm going to try and write to another drive and see if it freezes up to try and start the troubleshooting. Hey if it writes fine to another drive then I can start eliminating other possibilities... I did just buy one of those Supermirco rack-mount servers that is fully built. If my system was currently 17 data drives + parity I would move it over to the new hardware. I'm thinking about buying 3x 2TB drives to reduce my HD's to a total of 15 and moving it over to the new hardware...
November 17, 201015 yr Author Well I have tried to write to another disk and it still froze, though that was on the same card. I guess I will let it rebuild parity and try a drive that isn't on that card to see if the card could be the problem. FYI these two drives are connected directly to the card, no drive bays...
November 17, 201015 yr Well I have tried to write to another disk and it still froze, though that was on the same card. I guess I will let it rebuild parity and try a drive that isn't on that card to see if the card could be the problem. FYI these two drives are connected directly to the card, no drive bays... Another thing you can try is using a different NIC. Find/buy a cheapish Intel PCI NIC card and put it in the server, disable the onboard, and then try writing files to the server.
Archived
This topic is now archived and is closed to further replies.