How to remove bad drive during initialization population (no parity installed)

[Auggie]

I came across a possible bad drive during my initial population of the array, so I had not yet installed the parity drive.  Luckily, the drive was empty and I was just adding it to the array and formatting when UnRAID came across 2 errors.

 

The problem is that I can't "remove" the drive from the array as UnRAID keeps telling me it's "not present" and still lists the drive SN as part of the array even though I "unassigned" it on the devices tab.

 

Since I have not yet finished the array and therefore have not installed a parity drive, there is no "Restore" button.

 

What do I do now?

[Rajahal]

What version of unRAID are you using?  The 'restore' button was removed in version 4.5.4, so you'll never see it in that version (and presumably every future release as well).  The functionality is still available by running the command initconfig on either the system console or during a telnet session.  Just type initconfig and hit enter, the command will run instantly.  Refresh the web management page and unRAID will say 'Initial Configuration'.

 

I believe there is even a way to run initconfig through the web management page by typing in a particular URL, but I don't know the syntax.

 

Initconfig should solve your problem.  It will make unRAID forget about all disk assignments and it will immediately invalidate parity and initiate a new parity sync if you have a parity disk installed.  It will not erase any data.

 

Also, are you sure the drive is bad?  The errors could have been caused by a loose connection or something similar (and easy to fix).  Post a syslog and other hero members can advise you on the cause of the errors.

[Auggie]

What version of unRAID are you using?  The 'restore' button was removed in version 4.5.4, so you'll never see it in that version (and presumably every future release as well).  The functionality is still available by running the command initconfig on either the system console or during a telnet session.  Just type initconfig and hit enter, the command will run instantly.  Refresh the web management page and unRAID will say 'Initial Configuration'.

 

I believe there is even a way to run initconfig through the web management page by typing in a particular URL, but I don't know the syntax.

 

Initconfig should solve your problem.  It will make unRAID forget about all disk assignments and it will immediately invalidate parity and initiate a new parity sync if you have a parity disk installed.  It will not erase any data.

 

Also, are you sure the drive is bad?  The errors could have been caused by a loose connection or something similar (and easy to fix).  Post a syslog and other hero members can advise you on the cause of the errors.

Ah, okay, I'm running the latest 4.5.4!  Before reading this post, I simply put in my BASIC flash drive, detached all other hard drives, booted up and am doing a SMARTCTL on it as I right this.  It just finished with no errors.

 

I'm now about to run the preclear script to see if this fixes it...

 

EDIT: I checked the syslog after boot and didn't find any errors shown regarding the drive in question.  The error I got was from web GUI when I tried formatting it, when ready it showed it still unformatted with 2 errors.  Do I check the syslog after that for the type of errors encountered during the formatting?  I really don't want to wait through a day of "pre-clearing" this possible-easy-to-fix drive.  FYI: it was formally an -EARS HFS+ formatted drive that I added to the array, formatted via UnRAID, then realized I didn't jumper 7-8, so shut down, jumpered, formatted under OS X as HFS+ (in an attempt to force UnRAID to show it as needing to be reformatted) then re-installed into the box.  Now I'm getting the 2 errors...  All my other -EARS drive I jumpered prior to adding to the array and formatting.

[Rajahal]

The syslog is cleared when you reboot the server, so any previous errors won't show up there.  In the future, always try to capture and save a syslog before rebooting.

 

If you are currently seeing 2 errors, then capture a syslog and post it here.  It sounds like the errors may be related to the jumpering of the drive, but there's no way to know without looking at the syslog.  It could still be a loose cable or an overtaxed PSU or a number of other things.

[Auggie]

It could still be a loose cable or an overtaxed PSU or a number of other things.

 

No, not a loose cable as I changed power and SATA plug from another port (using the cables from the other port) and it still got errors.

 

NOW the WEIRD thing.  I thought perhaps when I reformatted under OS X that I did a GUID partition map vs Apple Partition Map (which I had done to all my other cross-platform usage drives).  So I did another reformat to APM HFS+, put it back into the UnRAID box, re-added it under the device tab and try to start the array... but it waits a little then dumps me back into the START button, even though it says the configuration is valid.

 

Huh?

 

EDIT: OKAY SUCCESS!  But after freakin' from the above, I ran the INITCONFIG, unassigned the questionable drive, and was able to start the array.  Whew!  Next step, re-add the drive to the array and it shows up unformatted!  Yippee!  The final test, hit the format button, sacrifice a chicken feather, cross my fingers... Viola!  No errors and full array is up!  I'm sure this drive is okay as its a couple months old and had no previous problems, and I did things incorrectly when adding this drive unlike the previous 4 drives.  I have one last drive to transfer, so after verifying the copied files to this, I will finally put in the parity and be more at peace once that is fully built.

[Rajahal]

Great, sounds like you're all sorted then.

[Auggie]

Great, sounds like you're all sorted then.

Yea, all transfers complete, parity drive installed and building parity-sync as I type.  The only thing left is to await my AOC-SASLP-MV8 board so I can install the final 7th empty drive and I can button 'er up!