This is my first time posting so sorry for any missing info.  

 

 

I had one 12 TB drive showing read errors and went offline.  I shutdown and changed the cable on the drive and was able to bring it back online.  I ordered a replacement drive just in case. It lasted a day and continued to have read errors so I switched some other cables and best i can tell the one HBA cable was having issues. So i was able to move all drives off that cable.  But in the process of switching those cables i messed up the power connector on a 8 TB  drive that had the 3.3v line taped.  Which gets me to the last state which was the 8 TB offline in emulation mode and the 12 TB listed as unmountable. 

I took a picture of the last array setting (to get the drive info) before i removed the drive 3 and 5.  But did not take a final diagnostics at the time.  I did take one a few days before when the first drives started giving errors.  I have attached both.

At this point i shutdown the system re-taped the power on the drive and connected the two drives with errors to an external dock and was able to copy off some files as an extra backup.  I can access the drives fine in the external dock but copying is very slow with the 12TB so i assume it needs to be replaced.

I'm hoping the 8 TB is ok but may be out of sync with parity as it was in emulation mode.  I have a 14TB to replace the 12TB. This is my first time dealing with multiple failures and understand i have made the issue worse by trying to half fix it from what i could read online.   

At this point i was hoping to get some advise on the best route forward at this point.  This is a primarily a media server so i would like to restore what i can but everything can be replaced.  I assume the advise would be to bring it online and run a current diagnostic.   Could you please suggest the best steps to get to that ie connect all drive, replace the 12 with tht 14, etc.

Any advise would be greatly appreciate.  Thanks.

tower-diagnostics-20250320-1449.zip

Quick reply to you.  First and most important!!!  DON'T panic.  Start taking a few deep breaths.  Then wait until you get a real Guru to step up.  They are all over the world and time differentials may mean it may tomorrow before they see your post.

 

Second thing.  I doubt if you have lost any data at this point.  Remember that your data disks are formatted with standard Linux formatting schemes and data recovery tools are available for all of them.  What ever you do, DON'T format any disk with the thought that that is necessary to recover data.  It  is not!!!!

 

Third thing.  I would just set that new 14TB drive aside for a while.  Let's get those disks back in service and see what the damage is. (if any...)  After everything is working properly, then do the upgrade.

 

Fourth thing.  Put the tape (Kapton is preferred) on the 3.3V pin on the drive!    That way, it will always be in the best location to do its job.  (I have a drive with tape on it.) 

Really wish you had asked for help before taking the disks out of your server. You may have to New Config to get your array setup again, which will make it difficult to rebuild anything.

 

36 minutes ago, Bartok27 said:

bring it online and run a current diagnostic

See if it will even let you start the array if you put all disks back exactly as they were. Then post new diagnostics.

Thanks for the quick reply.  I have reconnected the drives and the array came online with disk 5 disabled.

 

tower-diagnostics-20250404-0959.zip

 

Let me know what you think and what i should try next.

Disk3 mounted now, but there appears to be a cable issue, replace the SATA cable and post new diags after array start.

I switched drive 3 to a new sata.   It was connected to one an one of the connectors on an HBA card and is now connected directly to an onboard sata connector.  It appears to still be failing.   I have two more new sata cables and i new HBA cable I could try.  I'm not sure how to best determine if its the cable, the connector, HBA card port or drive thats the issue. let me know what you would suggest next.  Thank you for the help.

 

tower-diagnostics-20250404-1320.zip

It still looks more like a power/connection issue, did you also replace/swap the power cable? Also, run an extended SMART test.

I swapped the power connector. Still same error,  so i tried the new power connector and the HBA cable connector and it looks better.   

I have started an extended smart test so far its just been sitting at 10%.   Will let you know when it finishes.tower-diagnostics-20250404-1427.zip

Not sure if this is an issue for you but I found my system had an overloaded SATA power cable.  I think I had 4 or 5 drives on one cable and the system was unstable.  When I put another power cable into the PSU to share the load, everything got better.  The other thing I did was to keep the drives spinning 24/7.   I have read conflicting accounts as to whether that lowers the life of the drive; however, the array has been rock solid since I made these two changes.  I am beginning the suspect that a lot of the cable problems are not SATA connectors but power connectors on overburdened PSU cables. 

22 minutes ago, demanding-chief3698 said:

I put another power cable into the PSU to share the load, everything got better.  The other thing I did was to keep the drives spinning 24/7

Thanks for the suggestion, i will have a look at that also.

 

1 hour ago, demanding-chief3698 said:

The other thing I did was to keep the drives spinning 24/7.   I have read conflicting accounts as to whether that lowers the life of the drive; however, the array has been rock solid since I made these two changes. 

 

I suspect that the overloaded connector was your real problem.  Personally, I avoid using SATA-to-SATA power splitters.   I only use the Molex-to-SATA ones.  I, personally don't think that the SATA connector design has decent power handling capacity to be used for multiple drives. 

 

Regarding whether powering down lengthens drive life, there is no data one way or the other. But it definitely lowers power consumption for most Unraid users.  When I watch a movie, there is only one drive spun up.  When I not watching a movie, all the drives are spun down 95%+ of the time!  And most of that 5% is the time required for the monthly parity check. 

It been about 21 hours since i started the extended smart check and it still says 10% complete.  From a bit of reading its unclear if this may just be a display thing and its actually running in the background. The spin down delay on my drives are set to default.  Which i from what i can tell is never spin down so dont think that is the issue.  Im leaning toward just waiting another day and see what happens but just checking in to see if there is anyway to tell if its actually still running.

Thanks

 

post smart report for that disk, or diagnostics includes smart for all disks.

here is the current diagnosticstower-diagnostics-20250405-1215.zip

1 hour ago, Bartok27 said:

It been about 21 hours since i started the extended smart check and it still says 10% complete.  From a bit of reading its unclear if this may just be a display thing and its actually running in the background.

The progress updates at 10% intervals and the check should take something like 1.5 hours per TB in my experience.  
 

Thought I had already posted this, but apparently I didn't Submit.

 

If we're still talking about disk3 last self-test was short and completed.

SMART Extended Self-test Log Version: 1 (1 sectors)
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Short offline       Completed without error       00%     32611         -
# 2  Short offline       Aborted by host               90%     32520         -
# 3  Extended offline    Aborted by host               90%     32269         -

 

I ran a short self test first just to see if it would run.  Then i ran the extended test.   The extended appears to not have finished. Last i check it was still at 10% after 28 hours.  I left t running and today it shows the drive has spun down and 117,700 read errors on the main screen in unraid.  I tried spinning it up but it doesnt work.   I have attached the latest diagnostics.tower-diagnostics-20250406-1108.zip  Is there another way to force the drive to spin up to get the latest smart info?   Let me know what you would suggest to try next.  Thanks!

If disk3 cannot finish the test and keeps giving read errors it may be really failing, you can try cloning it with ddrescue, the new disk should be the same capacity, if that works without many errors, you can then try using the clone to help rebuild disk5.

Disk3 appears to have disconnected. Not obviously a disk problem though. You can try to fix connection, or just replace the disk and see if that helps.

21 minutes ago, JorgeB said:

If disk3 cannot finish the test and keeps giving read errors it may be really failing, you can try cloning it with ddrescue, the new disk should be the same capacity, if that works without many errors, you can then try using the clone to help rebuild disk5.

Maybe I have lost track of the thread. but these latest diagnostics shows all disks mounted and each with plenty of data, but disk3 isn't responding and no SMART report for it.

 

Thank you for the suggestions.  I have a few questions about the two suggestions.  Basically either continue troubleshooting the connection or clone the drive

 

7 hours ago, trurl said:

Disk3 appears to have disconnected. Not obviously a disk problem though. You can try to fix connection

 

So you are suggesting to try other cables in case its still a connection issue? 

 

If i go the DDrescue route do i have to clone it to a 12 TB drive.  I do not have a spare 12 currently but i do have a 14tb.   

 

Also we havent look at the disk5 8TB drive at all.  Can it be added back?   If we were to add it back could i rebuild the 12 onto the new 14TB.

 

Sorry if thats a dumb option at this point we are way past my knowledge so just trying to learn and find the best route forward.

I just want to be sure of the options and route forward so i dont make things worse.

 

Thank again taking the time to help out.

 

7 hours ago, trurl said:

latest diagnostics shows all disks mounted and each with plenty of data, but disk3 isn't responding

If disk3 is the only problem you can just rebuild it to another disk same size or larger, but not larger than parity.

 

Stop the array, unassign disk3, start the array in normal (not maintenance) mode, and post new diagnostics.

I stopped the array and Disk3 disappeared and is listed as missing.  With the only options defaulting to no device and no unassign option. 

If i leave it as no device and try to start the array i get the error below in the logs.

 

I get Apr  6 22:21:42 Tower emhttpd: unRAID array: ERROR:TOO_MANY_MISSING_DISKS

 and the array will not start

 

I could shutdown and restart and see if it comes back?

Or i could shutdown change cables again and restart?

 

Let me know what you think

OK, I see now that disk5 was missing in those last diagnostics. But missing/emulated disk5 was mounted, so that means it was being emulated by all other disks including disk3. And disk3 was also shown as mounted, but it was really missing. Very confusing.

3 minutes ago, Bartok27 said:

shutdown change cables again and restart

Let's try that. We need disk3 so we can rebuild disk5.

I will shutdown and change the cables and post new diagnostics if it comes up.

 

Is there any way to add disk5 back to the array?  I think it was kicked out because it had a bad cable which was replaced previously

Yes i dont understand them both being listed as emulated.