Unraid 7.3.2. I can’t keep up with my Unraid problems. I go from one disk error situation to the next. I have five spinning drives (one for parity) and an SSD cache. Fairly regularly I get errors pointing to a bad cable or cable connection. I shut down the array, shut down the system, reseat the cable, start up, run SMART test, comes out good, rejoin and rebuild the array. Which is usually a two or three day process in total.

After the latest one on Disk2 (sdd) I replaced both internal SATA cables. Then I ran the extended (16 hrs) SMART test and it came out good. So I reinsert the disk into the array and start the rebuild and now I’m getting errors not only on that Disk2 but on another one as well. I’ve attached a file with the log for Disk2 below.

I’ve had this system for years and granted there have been many months and months of trouble-free operation, but then sometimes feel like I’m chasing false errors on and on. I was running Plex on it but got tired of it being unavailabe and moved it.

Could my controller be bad? Any ideas? 

Unraid homebuilt in a Rosewill tower case with

• Unraid 7.3.2

• Five drives, one 2 TB, two 8 TB, one 10 TB, one 12 TB, with the 12 TB as parity. Mostly older HGST Helium drives. 

• Z390 Aorus Elite

• i5-9500T

• 32 GB RAM

• Hard drive controller: LSI SAS2008 SAS/SATA PCI-Express card

• PNY Quadro 400

• One 1 TB Sandisk SSD caching drive

Unraid disk2 errors.rtf

descartes-diagnostics-20260806-1030.zip

Edited August 6Aug 6 by Willion

Please post the diagnostics.

4 minutes ago, JorgeB said:

Please post the diagnostics.

Done, they are now attached to my original post. Thank you.

They are not logged as a disk problem, and SMART looks OK. Cancel the rebuild, check/replace both cables for disk1, avoid any power splitters, try rebuilding again.

I've not seen power splitters mentioned before as a potential problem. I think I do have some splitters in there in there so I'll take a look and remove them if I can. The power supply is a new one from a year ago, a Thermaltake 80 Plus Smart500W, I'll see if it has enough connections without splitters.

I have seated and reseated those cables but I'll do it again.

Thank you.

Molex to 2 SATA splitters are generally OK, but avoid splitters with more than 2xSATA plugs, SATA to 2xSATA splitters are also typically OK, but never use more than 2 with SATA, not enough amps. Still, it's good to avoid any splitters whenever possible.

All is resolved now. I suspect that an underpowered PSU was causing CRC issues on one of the disks during the data rebuild. After replacing the PSU and two of the disks just stopped working completely and another was being flaky I remembered that three of my disks are susceptible to the 3.3V Pin issue and I had previously avoided that problem with some SATA to Molex to SATA cables with the old PSU. So, I then addressed the 3.3V Pin issue with a power cable mod. The array has been rebuilt and all is good now. Thank you for the help, much appreciated.