November 24, 200916 yr Hi All, I've done some research and can't find any potential solutions to my problem, hopefully it's not a simple error and I'm asking a n00b question. A little background. I'm running unRaid 4.4.2 and have been doing so for quite a while. It's been stable and just keeps chugging along. The config was 8 drives (& + parity) all Samsung 500GB. I needed some extra capacity and purchased 2 x 1TB drives on the weekend. The intention is to use one for the new parity, one as additional data, and remove one of the 500GB drives for use in my popcorn hour. In addition to the drive I also needed 2 more sata ports, so I got a generic PCI card as well. Here are the steps that were taken, all without any indication of problems. 1. Powered down the unRaid box. 2. Installed the sata card, and cabled it to the 3 in 2 cage. 3. Installed the 2 hard drives in spare slots in the 3 in 2 cage. 4. Powered on the box. 5. Stopped the array. 6. Swapped the parity drive for one of the new ones. 7. Started the array. 8. Allowed it to finish syncing. 9. Stopped the array. 10. Assigned the other 1TB drive as disk8 in the array. 11. The drive cleared no problems. At this stage everything was working properly. I copied some data onto the new drive, streamed some media from the drive. Also powered the box down and back up again with no problems. Monday night after work, I powered the machine on again to do a few things. 1. Copied more data from one disk to another via putty and mc. (I wanted to clear on of the old disk and remove it, rather than the old parity drive as the parity drive is harder to get at physically). 2. Uploaded some files to the new disk via FTP. 3. Was streaming a file to the popcorn hour when the popcorn hour stopped playing. 4. The putty session had an error relating to IRQ21. 5. The web interface had stopped working. Though I could create a new telnet session and do some things. (ls -l on the /mnt directory caused the machine to hang). 6. I tried to follow some shutdown command sequences (specifically) which hung on the sync command. 7. I tried a powerdown command which also seemed to hang. 8. Eventually I powered down by holding in the power button. (It was late and I need to shut the box off to sleep). Tonight I'm getting a red ball on the new data disk. The array is coming online OK, and I can browse around. OK, looking around after typing this all up it, I found a little more information. I think the disk is being disabled due to a write error cause by the IRQ problem? Is this feasible? If it is then I can just restore using the 'Trust my Parity' as per here and here? I've attached the syslog from the latest boot, I didn't get one when the issues were happening. I've also attached the out put from a smartctl. Thanks in advance for your help. mp
November 30, 200916 yr This syslog does not show any problems at all, so I am guessing you have now done the Trust My Array procedure, which would be correct. Your Disk 8 looks fine. I don't know what you saw exactly, but IRQ21 is only associated with the VIA disk controller, to which you connected the 2 new 1TB drives, Disk 8 and Parity. If something went wrong with that IRQ, then you could very well lose access to the VIA controller, and therefore lose access to both drives. That may have been what happened, and the Trust My Array procedure is perfect for recovering the array in that situation. If this happens again, then I would be suspicious of the new controller. Cheap is only cheap if it works correctly.
Archived
This topic is now archived and is closed to further replies.