February 15, 20251 yr Hello everyone, I just had a nice run of bad luck yesterday when the power failed. I do have an UPS on my drive shelf and on the Unraid server. There should be a script that checks the UPS and once the battery level is below a certain level, it should shut down the server. I'm not sure if that didn't happen and eventually when the server came back on one of my parity drives and one data drive have a red X next to it. What I have tried. - I did a complete shutdown of the array and re-seat the drives (both are viewable in the drop down for selection) then re-booted the server. When it came back, both drives were still down. Are there any other suggestions to try besides drive replacement? If we are headed to drive replacement, which one should be replaced first, the parity or the data drive? Thanks,
February 15, 20251 yr Community Expert 14 hours ago, mwoods98 said: I do have an UPS What UPS? Support for common UPS can be configured in Settings - UPS. There is also the NUT plugin. You should only run on UPS for very brief outage and configure it to shutdown for outage longer than very brief. Also you may need to configure some of the shutdown timers. See here:
February 15, 20251 yr Community Expert Since you rebooted before getting Diagnostics, can't see what happened before that. Emulated disk6 looks like it still has all of its data. Not very familiar with reading SMART reports for SAS drives. Do any show SMART warning (👎) on the DASHBOARD page? You can rebuild both at the same time. https://docs.unraid.net/unraid-os/manual/storage-management/#rebuilding-a-drive-onto-itself
February 15, 20251 yr Author 28 minutes ago, trurl said: Since you rebooted before getting Diagnostics, can't see what happened before that. Emulated disk6 looks like it still has all of its data. Not very familiar with reading SMART reports for SAS drives. Do any show SMART warning (👎) on the DASHBOARD page? You can rebuild both at the same time. https://docs.unraid.net/unraid-os/manual/storage-management/#rebuilding-a-drive-onto-itself Yes, unfortunately I did reboot, I'll try to remember to grab diags next time before rebooting. For now, it is emulated and looking at the drive and the settings it doesn't show any smart failures. However, I can't run any self-tests or anything.
February 15, 20251 yr Author 40 minutes ago, trurl said: What UPS? Support for common UPS can be configured in Settings - UPS. There is also the NUT plugin. You should only run on UPS for very brief outage and configure it to shutdown for outage longer than very brief. Also you may need to configure some of the shutdown timers. See here: Thanks, I'll review these when looking back at the logs NUT is not running and I can't verify that it was working before. I am going to assume its not as when looking at the logs I need to be on one level above to have this version of NUT running.
February 15, 20251 yr Author Update: I was able to get the drives to mount by creating a new configuration.
February 15, 20251 yr Community Expert @mwoods98 2 hours ago, mwoods98 said: Update: I was able to get the drives to mount by creating a new configuration. Did you let it rebuild both parity? You either needed to rebuild the disabled disks, or else New Config and rebuild both parity. The disks were disabled because they were out-of-sync. Rebuilding the disabled disks would be the normal thing to do since they were the disks out-of-sync. If you didn't rebuild anything then parity isn't valid.
February 15, 20251 yr Author 57 minutes ago, trurl said: @mwoods98 Did you let it rebuild both parity? You either needed to rebuild the disabled disks, or else New Config and rebuild both parity. The disks were disabled because they were out-of-sync. Rebuilding the disabled disks would be the normal thing to do since they were the disks out-of-sync. If you didn't rebuild anything then parity isn't valid. Yes, I started a parity rebuild.
February 16, 20251 yr Community Expert Why didn't you rebuild the disabled disks instead? It wouldn't have been any slower and 7 hours ago, trurl said: Rebuilding the disabled disks would be the normal thing to do since they were the disks out-of-sync. Unraid disables a disk when a write to it fails. The disabled disk is emulated by parity. The emulated disk can still be read and written. When the emulated disk is read, instead of reading the disabled disk, the emulated data is the result of the parity calculation from reading all the other disks. When the emulated disk is written, instead of writing the disabled disk, parity is updated just as if the disabled disk had been written. The initial failed write, and any subsequent writes, can be recovered by rebuilding the disabled disk. If you don't rebuild the disabled disk, those emulated writes are lost. It is even possible that some of the lost writes would mean the filesystem wasn't consistent anymore, and so would be corrupt.
February 17, 20251 yr Author 22 hours ago, trurl said: Why didn't you rebuild the disabled disks instead? It wouldn't have been any slower and Unraid disables a disk when a write to it fails. The disabled disk is emulated by parity. The emulated disk can still be read and written. When the emulated disk is read, instead of reading the disabled disk, the emulated data is the result of the parity calculation from reading all the other disks. When the emulated disk is written, instead of writing the disabled disk, parity is updated just as if the disabled disk had been written. The initial failed write, and any subsequent writes, can be recovered by rebuilding the disabled disk. If you don't rebuild the disabled disk, those emulated writes are lost. It is even possible that some of the lost writes would mean the filesystem wasn't consistent anymore, and so would be corrupt. The person who I was working with recommended this method.
February 22, 20251 yr Author On 2/17/2025 at 1:05 PM, trurl said: Maybe you should have gotten a second opinion Post follow up: The array is back, no data loss, both drives are back with the parity intact. @trurl While I appreciate your assistance, there is often multiple ways to solve a problem. Respectfully, thanks.
February 22, 20251 yr Community Expert Sometimes a user will have a disabled drive that had a lot of files written to the emulated disk. By not rebuilding the data disk, all those emulated writes would be lost. I recall someone who was missing about a week's worth of data because of that very thing.
February 24, 20251 yr Author Agreed, and this very well could be an issue. Once I got things back, I made sure to make sure there no writes to the array so that wouldn't be an issue.
February 24, 20251 yr Community Expert Do you have Notifications setup to alert you immediately by email or other agent as soon as a problem is detected?
March 8, 20251 yr Author On 2/24/2025 at 1:09 PM, trurl said: Do you have Notifications setup to alert you immediately by email or other agent as soon as a problem is detected? Sorry been a bit busy but wanted to get back to this when I had some free time. I do not have notifications setup when there is a power failure. If you have a link or a guide that you would suggest, please post it here and I'll be glad to set it up. Thanks
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.