January 9, 201610 yr Hello, I am running 5.0.6 with 9 data drives, 1 parity drive and 1 cache drive. Drive in question: Disk 7, /dev/sdh, referred to in the question below as 'bad'. TL;DR;: I had a drive removed from the array after a failed parity check, ordered a new one, ran the smart report and think the bad drive is okay. What should I do next? About a week ago, I had a power outage and my UPS didn't last as long as the power outage. After the power came back, I rebooted my server and initiated a parity check. During the parity check, there were some read errors (I did not capture the log), and the bad drive was pulled out of the array. I started the parity check before I went to bed and when I awoke, it had failed. Not having used my unRaid server interactively in quite a while, I immediately ordered a new drive to have it on hand when I delved deeper into the problem. The drive arrived yesterday and I hit the wiki's and faqs to determine the next steps. 1. I captured the syslog files before rebooting. 2. I rebooted. 3. I ran the short smartctl test and it passed. I have attached the syslog for before the reboot and after the reboot and the smart report for the short test. The smartctl test is provided in full. I had to trim the syslogs to perceivably the most relevant parts due to file size. I am looking for the best guidance moving forward. I'm inclined, based on the wiki, the smart report and the post-reboot syslog to: 1. Re-enable the drive 2. Reconstruct it, based on the instructions here: http://lime-technology.com/wiki/index.php/Troubleshooting#Re-enable_the_drive 3. Shutdown the system. 4. Install new drive, pre-clear it, and wait for a failure when I need it. Thank you for reading my post and for any assistance you can provide! -Chris syslog-before-reboot.snippet.txt syslog-after-reboot.snippet.txt smart.report.20160109-105210.dev.sdh.txt
January 13, 201610 yr Author I was hoping someone would confirm or reject I'm on the right track and would be doing something that sounds correct, before I do something that causes more problems. Any help is appreciated, please!
January 13, 201610 yr Sounds good, I would probably do an extended SMART test to confirm disk is good before rebuilding.
January 13, 201610 yr Author Thank you for the confirmation and notes to run the long test. It's running now; fingers crossed.
January 15, 201610 yr Author I have enabled the drive and reconstructed it. The result is valid parity. Would you recommend I run a new parity check before using the array?
January 15, 201610 yr Parity check is always a good option to boost your confidence in the array, but If both the extended SMART test and disk rebuild completed successfully I would have no problem using it in the meantime.
January 15, 201610 yr Author Alright. I'll run it for good measure and enable a scheduled parity check from here on out. Thank you and have a great weekend.
Archived
This topic is now archived and is closed to further replies.