Hi all,

 

Looking for guidance on best steps to proceed.  

 

Over past couple of months, I noticed one of my drives having a very slowly creeping UDMA CRC error count.  For the drive in question (Disk 3, WD80EZAZ), things had seemed to stabilized around a UDMA CRC error count of 29 for sometime.  The drive in question is a shucked drive from a WD elements enclosure using Kapton tape over the 3rd pin trick, but working flawlessly for the past 5 years or so.  Last weekend I had time to open the box and take a good look at things and double check connections as well as the Kapton tape.   Everything appeared to be ok and so just double checked connections.  Been monitoring every day and the count hasn't moved.    Since I also had the box open last week, I upgraded some ram, installed a new HBA card, and introduced another drive (also shucked).  Not sure it's relevant, but providing in the outside chance the context matters.  Everything running smooth and stable for rest of the week.

 

Fast forward to yesterday morning where I found out the drive (Disk 3) had been disabled, seemingly as a result of multiple failures to write.   I opened up the box.  Inspected the drive and tape again and decided to swap out the corresponding sata cable.   I have run a short SMART test (passed) and an extended test (also passed, results attached).  Drive is showing as healthy (with the exception of the [still only] 29 UDMA CRC errors).  I've reviewed all the threads on the same topic and I THINK it might be ok to rebuild my drive upon itself, but I wanted to seek some guidance first.  I'm ok to swap the drive with the new one if I have to since it is also been running for 5 years.  Since taking the array down yesterday when I noticed the issue, I have yet to try and start up the array again.  

 

My questions:

1) Given the diagnostics and extended test results, would you recommend a rebuild upon self?

2) I've seen comments in other threads, like in this one (https://forums.unraid.net/topic/136397-device-is-disabled-contents-emulated-what-is-my-next-move/ ), that it would be better to maybe replace with a new drive instead of rebuild in case something goes wrong (see Itimpi's comments on July 22, 2023).  Let's say the rebuild of a new drive doesn't work, how does having the old drive help?  Maybe I can still try and get the contents off safely if I put the drive up and things are up and running and emulated?  

3) Since I'm more fussed about the data/downtime than the cost, I'm considering as a first option to just replace the drive with a new one and rebuilding and monitoring.  Then monitoring/taking a better look at the drive in question and re-purposing (cold storage, backup, or re-introduced as another drive into the array) if signs continue to point to external factors instead of the drive.

 

Thanks in advance for any guidance, suggestions, or clarifications you can provide.  

 

 

 

 

 

tower-diagnostics-20250419-0700.zip tower-smart-20250418-2332.zip

1 hour ago, GotDaFunk said:

I have yet to try and start up the array again. 

This means we are missing one critical piece of information. Is the emulated disk mounted?

 

Start the array in normal (not maintenance) mode and post new diagnostics.

Thanks for the quick reply.  Started it back up, and I can see it says :

Device is Disabled, Contents Emulated

 

So I believe so.  Device shows up in the array as disabled.  Browsing around the shared folder structure and things appear to look ok, files wise.  Updated diagnostics attached.

 

tower-diagnostics-20250419-0917.zip

Yes, emulated disk3 is mounted. Should be OK to rebuild on top.

 

As you may already know, UDMA CRC are connection problems.

 

All plugs should sit squarely and firmly on the connector, with no tension in the cables. Don't bundle data cables to make things "neat". Be sure to also consider power connections. Ideally no more than 4 drives per PSU cable. Are you using any power splitters?

3 hours ago, GotDaFunk said:

Kapton tape over the 3rd pin trick

Another way to tackle that problem is with a 4-pin Molex cable from the PSU then a Molex to SATA adapter. There is no 3.3V in the 4-pin.

Thanks for the replies.  Much appreciated!
 

I might still put in a new disk.  I’ve also read about the 4 pin molex, thanks. I do have a large roll or Kapton tape to make use of, but it certainly takes the hassle out of it. Ironically I needed that Kapton tape for my HBA card install last week after it didn’t auto detect when installed. 

 

Any insight into item 2)?  So my drive can currently be emulated.  If instead of trying to rebuild on top, I try and repair my array with a new drive, and something goes wrong, can I stick my current drive (emulated) and then maybe I have the option to get stuff out ?  Maybe using that plugin called unbalanced (or whatever it’s called) to try and get everything that might be written on that drive off of it? 
 

thanks again!  
 

 

 

 

 

Edited April 19, 20251 yr by GotDaFunk

4 hours ago, GotDaFunk said:

Any insight into item 2)?  So my drive can currently be emulated.  If instead of trying to rebuild on top, I try and repair my array with a new drive, and something goes wrong, can I stick my current drive (emulated) and then maybe I have the option to get stuff out ?  Maybe using that plugin called unbalanced (or whatever it’s called) to try and get everything that might be written on that drive off of it? 

Any time a disk is disabled or missing, Unraid emulates it from parity plus all other disks. Emulated doesn't actually refer to any physical disk, it is just the contents of the disk slot as calculated from parity.

 

Usually, when we talk about "repair", we mean repairing a corrupt filesystem. Any time a disk that should have data on it is "unmountable", it needs filesystem repair. What you probably mean by "repair" should be referred to as rebuild. A disk is mounted when the OS loads its filesystem to allow access.

 

Rebuilding to a different disk does leave more options, but those options aren't usually needed when the emulated disk is mountable. If the emulated disk is unmountable, the results of attempting filesystem repair may not be as good as the contents of the original disk.

 

The only way to return the original disk after attempting a rebuild to a different disk would be to New Config with the original instead of the replacement, and rebuild parity.