Every time I restart  the server, I loose drive assignments.  When this happens, the lost assignments are not necessarily in an pattern that I have seen.  This last time, we had to kill to power to the house, so I did a Stop and then powered down the box.  After the power was restored, it came up showing missing drives.  When I went to Devices, they were there so I could assign them, except for the parity drive.  I tired Restore but no luck, so I shut it down again and brought it back up.  Again drives were shown as not assigned so I assigned them again, including the parity (this time it was there).  It is currently going thru the 18 hour sync (yes I have 14 1TB drives so it takes a while) and everything appears to be there, but if there is a power failure and I am not there, anyone wanting to use the server is SOL due to this having to assign the drives again.  Any ideas or suggestions?  I need to solve this so if the power goes out and I am not home, anyone can just turn on the server and get back to running.

We need to see your syslog, please see my sig for the Troubleshooting link, for help with capturing a syslog.  (The unRAID Wiki has a lot of helpful information too.)

 

Actually it would be very helpful if you could provide us with 2 syslogs, one when the drives are there, and another when they are missing.

 

Currently, forum attachments are not working, so you will need to use an online file upload place, such as:

• http://pastebin.com/
  
• http://www.mediafire.com/
  

 

To save you a LOT of time, try the Trust My Array procedure, after you have re-assigned the missing drives.  You can abort the parity check that will start, but let it go 2% or 3% first.

I guess it does not happen all the time.  For example, I did the syslog steps this morning  which does start about the time I was bringing it back up, but that was probably after I had done the reassignments, and when I brought it back up, all was there.  The next time it does it, I will get the syslog and post it.

Some motherboards have an annoying characteristic of shuffling the disk assignments on every boot.  This results in unRAID giving out different device assignments.  Sometimes unRAID can sort through the muck and still correctly assign drives to slots on the device page, other times not.  I have an older motherboard (now relegated to my test array) that had this annoying tendency, and required manual fixup from time to time.  My array tends to run 24x7 so this was never a huge issue.  (A UPS also helps)

 

What concerns me more is the fact that the parity disk was missing.  This should not happen!  My newest motherboard is a very quick booter - and I had a problem with the first drive or two not getting recognized sometimes.  To fix, I turned off quick booting, so the MB does a short memory test at each boot.  This gives the drives a chance to come up to speed and have not had a recurrence since I did that.

 

The other thought is that maybe your PSU is not supplying enough amps at boot (the most strenuous time for a PSU).

 

Finally, pressing the RESTORE button is something you should NOT do, unless you are setting up an array for the first time, or as part of using the "Trust my parity" procedure.  Pressing it will result in a complete loss of parity protection and inability to recover from a drive failure.  DO NOT PRESS IT unless you have confirmed here that it is the right thing to do.  It is better to not start the array at all than having to press the RESTORE button to start it.

Had storms come thru last night so I powered everything down in advance and unpluged.  When I brought it up again, half the drives were unassigned again, including the parity.  I made a copy of the log, assigned all the drives except the parity, then shut it down again.  Then I brought it back up and parity was there and did not even have to assign it.  Not sure how to post the log except to copy and past.  Is that the way?

Had storms come thru last night so I powered everything down in advance and unpluged.  When I brought it up again, half the drives were unassigned again, including the parity.  I made a copy of the log, assigned all the drives except the parity, then shut it down again.  Then I brought it back up and parity was there and did not even have to assign it.  Not sure how to post the log except to copy and past.  Is that the way?

 

Apologies for the delay, I somehow missed your note from a few days ago.

 

Why would you have corrected all drives except parity?  Did you start the array that way?  (Hopefully not!)  If your device assignments come up scrambled, just fix them and start the array.  No harm no foul.

 

Again, the reason for the device scrambling is that the motherboard is not consistently assigning drives in the same order.  My motherboard that did this would tend to keep them in the same order for a while, and then something would happen and they'd be in a different order for a while.  As I said, once the server is up and running, it doesn't matter until you have to reboot.

My problem is if there is a power issue which causes a need to reboot AND I AM NOT THERE.  The others are not techie enough to get the assignments right.  That is why I am trying to make sure it works every time. 

 

Can I just copy and paste my log in here?  And is there a way to get the MB to do the assignment the same every time to avoid this?

I don't know if you saw my earlier post?  We would be happy to help, but we still need to see both syslogs.

I don't know if you saw my earlier post?  We would be happy to help, but we still need to see both syslogs.

 

Yes I did but not sure how to get to you.  That is why I asked if I should just copy and paste, since I have no idea what those other two things are that were listed when saying the attachments are not working.