CtznSnips

Do I need to rebuild the disks or is there a quicker way to restore them? Also, this seems to happen to me fairly often, every 6-8 months or so. I'm wondering if I should replace the power supply. Does anyone have recommendations for a power supply? One that can handle 11 HDDs and 4SSDs

Logged onto the server today and saw 3 disks with errors, 2 of which have been disabled. Nothing unusual was happening when the errors occurred. Help? yoda-diagnostics-20240430-1639.zip

Thanks for your help. Turns out there were 10 drives powered by the same PSU cable (with 3 SATA power connecters) using splitters. I guess when I built this server 4 years ago I didn't know any better. I spread the 10 drives across more cables and removed some splitters. So far the rebuild is at 18% without issues. Hopefully this fixes it for good. Is there a general rule for splitting SATA power and how many drives you can power per connection directly from the PSU? Maybe I should look into a new power supply...? I have 8 drives being powered now on 6 SATA power connectors direct from the PSU.

yoda-diagnostics-20230830-2029.zip This seems to happen to me every 6-8 months - out of nowhere I get disk errors but the disks themselves seem fine. Usually I can resolve it by re-seating cables. Yesterday I got errors on a couple of disks, I tried to re-seat cables and start the array back up but I continued to get errors. I had a back-up HBA card and cables so I tried replacing those which worked while I restored one disk. The restore completed successfully, so I started to rebuild parity, then again I started getting errors. I'm running out of ideas here, any help is appreciated.

Thanks again. I reseated the cables and the parity-sync is going well with 16 hours or so to go. The majority of my hard drives are connected via LSI HBA card in IT mode with SAS to SATA cables. Do these go bad often/frequently, and would it be a good idea to have a backup set of cables? Thanks

Thank you for the reply. Forgive my ignorance here - can I re-seat the cables while the server is running? Or is the proper procedure to shut down, re-seat, and start the server back up? Also, how do I re-sync parity? Do I need to stop and start the array?

Version: 6.11.0 Sometime after my normally scheduled parity check began on 3/7 my first parity drive (sdb) reported some read errors and was disabled by unraid. The parity check completed without any errors. I'm hoping for some help in determining if the drive is still ok. It's a 14TB WD elements drive that I shucked and less than 1.5 years old. In the meantime I do have another 14TB drive unassigned (sdc). Should I rebuild parity on the same drive or flip with the other drive? What's the proper way to do so? And either way, is that drive which is reporting errors still ok? I just completed an extended SMART test on the drive in question. Diagnostics attached yoda-diagnostics-20230309-1302.zip

Yes, I checked every other device on the network.

Yes I do, but I'm not sure how to bypass the gpu. I only have a vga monitor to plug into the motherboard. I guess I could pull the graphics card but I'd need to power down first

I hope someone can help me diagnose what is going on here, I don't understand how something like this is even possible. This is the second time this has happened (it last happened exactly 1 week ago overnight Friday into Saturday morning). I'm running unraid 6.8rc-5 on my server. It works fine all week and then some time overnight Friday the server disconnects from the network (not accessible by browser using the server name or the IP address and shows as not connected in the Google wifi router app.) However, on my Windows PC, I can still access shares and play video files hosted on the server. The server is assigned a static IP address by Google wifi. Last week it resolved after a server reboot. Anyone seen something like this before and know what might be wrong?