So

 

1. put all of the disks that have data in the enclosure

2. start up unraid

 

i know there is a way to tell unraid to ignore the current configuration and just build a new array.  that should take all the data on the disks and build a new parity disk from them, and when it's done I will have a new array minus whatever data was on the disks I included from the array.  Correct?

 

So

 

1. put all of the disks that have data in the enclosure

2. start up unraid

 

i know there is a way to tell unraid to ignore the current configuration and just build a new array.  that should take all the data on the disks and build a new parity disk from them, and when it's done I will have a new array minus whatever data was on the disks I included from the array.  Correct?

This wording has me a little confused.

Sorry.

 

I'm trying to confirm the process to create a new array with disks that already have data on them. I've never done this before. I always started with a couple of empty disks and kept adding on.

 

Sent from my SM-N920V using Tapatalk

 

 

So

 

1. put all of the disks that have data in the enclosure

2. start up unraid

 

i know there is a way to tell unraid to ignore the current configuration and just build a new array.  that should take all the data on the disks and build a new parity disk from them, and when it's done I will have a new array minus whatever data was on the disks I included from the array.  Correct?

I agree with all that but still don't know what the struck part means. It is the 'minus" in combination with the "included" that have me scratching my head.

Very sorry.  That was a typo.

 

I was just trying to say that if i rebuild an array without disks 3 and 5, the array would be perfectly fine when the build is done, it would just not have any data that was on disk 3 and 5.  I am trying to confirm that the data on the rest of the disks would be in tact and parity would be functioning properly.

Very sorry.  That was a typo.

 

I was just trying to say that if i rebuild an array without disks 3 and 5, the array would be perfectly fine when the build is done, it would just not have any data that was on disk 3 and 5.  I am trying to confirm that the data on the rest of the disks would be in tact and parity would be functioning properly.

Consider it confirmed. If you have any questions when you try to do this, be sure to ask.

You haven't answered the question r.e. whether you have one or two parity disks.  As already noted, if you only have one parity disk, you can't rebuild anything if there are two failed disks.

 

As for a disk that freezes the system --> I've seen this on several systems ... both Windows & Linux.  It's generally caused by a fault in the disk's controller board that causes the controller to freeze, but can also be the result of a faulty port on your disk controller or a short in the cable.  Have you tried using a different SATA cable and/or a different port?

 

I also have to ask ...

... Redball was for several weeks.  Maybe a month.

 

... WHY would you run a system with a redballed disk for this long !!??    I wouldn't run it an hour ... let alone a month !!

 

 

 

Very sorry.  That was a typo.

 

I was just trying to say that if i rebuild an array without disks 3 and 5, the array would be perfectly fine when the build is done, it would just not have any data that was on disk 3 and 5.  I am trying to confirm that the data on the rest of the disks would be in tact and parity would be functioning properly.

 

Note that if you have dual parity, you do NOT have to build a new array excluding your old disks 3 & 5 => you can simply REPLACE both of those disks with your current configuration and they will both be rebuild with all data intact.

 

If you only have single parity now; then indeed you can simply do a new config with new disks in place of the two failed ones; then let the system generate parity (so you'll now be protected); and then you can attack the issue of trying to recover data from the two failed disks.

 

... You may want to consider using TWO parity disks with your new configuration, so you'll be better protected in the future.

 

I know, I know.

 

1 Parity Drive.  I may indeed add a second when my new system is up and running.

 

I didn't have notifications, and the red ball was not obvious at first. The disk has been full for a long time and I just had a few errors, then more, then finally a red ball.

 

Bottom line I just got complacent.

 

I won't again.

 

Sent from my SM-N920V using Tapatalk

 

 

I think you're the 3rd person in the last week that didn't have Notifications setup (4th if you count the guy still running v5) and then accidentally discovered that things had gone past the point of no return. One of the problems I have had is keeping you guys threads straight. I have to read them all again each time just to remember who I'm talking to. I consider Notifications the most critical improvement in V6.

 

garycase, I had not really spelled it out, but I could see he had no parity2 due to slot29 empty which I included in the drive list I posted on page 2.

... I consider Notifications the most critical improvement in V6.

 

For v6 I'd agree ... but at the 6.2 point it's a tough call between notifications and dual parity 

 

... although I guess I'd agree with notifications even then => if somebody isn't going to notice when a drive fails; they may also miss the 2nd drive failure as well.  On the other hand, it depends on how notifications are configured;  if they're getting a lot of notices and warnings they may routinely dismiss (delete) them, and not notice the one important alert that indicates a failed drive.    e.g. neither option is really foolproof.

 

I would say the importance of dual parity increases with drive count, but you still need to be vigilant with or without dual parity and notifications makes that very easy.

 

Seems some just run for weeks on end never even knowing one of their drives is being emulated, thus increasing the wear of all the other drives.

I got unraid to start with that bad disk in a USB enclosure in a new PC.  I am pretty sure that disk is shot.  Attached are the new diagnostics.

tower-diagnostics-20160909-1813.zip

I got unraid to start with that bad disk in a USB enclosure in a new PC.  I am pretty sure that disk is shot.  Attached are the new diagnostics.

Could you elaborate on that first sentence a little bit? Are you saying unRAID is in a new PC, or just the bad disk? Or have you booted another install of unRAID on a different PC just so you could have a go at the bad disk?

 

What exactly is the plan you are proceeding with, and what actions have you taken? I'll look at the diagnostics later when I don't have to guess what has happened since the last ones.

Sure.  I pulled out an old PC.  I put the latest unraid RC4 on a flash drive and i booted it up.  The bad disk is connected to this PC in that cloning USB enclosure I mentioned.  So my plan was to try and clone the bad disk.  But first I wanted to see if this old pc could read the bad disk.  I was hoping UNRAID would not just hang like my actual unraid server. 

 

Now even tho this is an old PC, its actually a much much newer pc (core i5 650) than the ancient one I use for unraid (Dual AMD Opteron 280) which is why I bit the bullet to upgrade in the first place to the dual Xeon L5520s. (http://www.cpubenchmark.net/compare.php?cmp%5B%5D=522&cmp%5B%5D=767&cmp%5B%5D=1260).  Anyway, I thought maybe the newer PC could handle whatever errors this drive was throwing better than my current unraid system.  It seems it did because I can at least get it to boot.

 

Also, for what its worth... I have a diagnostic from a few weeks ago, before I started mucking around.  I can't really make heads or tails of the smart report, but it contains report for the two bad drives...

 

The redball: Z1FPLM3

The 2nd bad drive: W1F0ZQ9M

 

Update... I can mount and see the files on W1F0ZQ9M so I am going to attempt to clone it onto the new disk 3 TB disk I just purchased.

 

 

megatron-diagnostics-20160822-0735.zip

I got unraid to start with that bad disk in a USB enclosure in a new PC.  I am pretty sure that disk is shot.  Attached are the new diagnostics.

 

There's no SMART for the problem disk, USB enclosure usually don't pass SMART info, but if the disk works you can use DD to clone it on that server.

Update both disks have problems.  I could see some of the data, but I also got lots of I/O errors eventually and I also couldn't clone it.  It's going to have to go in the bin.

 

If you don't have backups of your data, you might at least try to copy some of the files => you may be able to recover at least some of the data.

 

I presume when you rebuild the system you'll be sure to enable notifications 

... and perhaps use dual parity as well.

 

(But nothing substitutes for having good backups)

You can try to copy what you can from the bad disk like gary said. And of course you still have the redballed disk to try to recover files from. It depends on how much work you want to go to and how important the missing files are and which disk you think is most important.

 

There isn't going to be any way to rebuild the redball without the missing disk, but you still may be able to access some of the files from the redball physical disk if we make unRAID accept it again, and you might be able to repair its filesystem to get even more from it.

 

And if you think it is worth the effort we could try to rebuild the bad disk to a new disk using the badly out-of-sync redballed disk. We would make unRAID accept the redball and instead rebuild the bad disk and maybe try to fix its filesystem.

 

All this would be complicated and probably would be far from perfect and might not even work, or would leave you with a lot of "lost+found" files that were recovered but no idea what folder they belonged to or what they were named.

 

If you think it is worth the effort maybe we can come up with a plan.

I have been converting my drives to XFS and copying data from reiserFS drives like I see explained in the sticky.

 

When I went rebuild a new array, I accidentally added one reiserFS drive to the array.  Unraid told me it all of the XFS drives have no FS and I need to reformat them.  I said no and canceled of course.  I realize now that I should have left the reiferFS drive out of the array or just did one more conversion first.

 

Now my unassigned devices look like the attached.  I did not reformat them, but unraid does not recognize the XFS disks. 

Please tell me there is some way to restore XFS and all of the files on these disks.

 

Thanks.

 

FYI -- I also posted this in the XFS conversion thread, not sure where it belongs.  I am happy to remove one of these posts. But the issue started here.