[zfs-discuss] !HELP! Lost my pool?!?

Manuel Amador (Rudd-O) rudd-o at rudd-o.com
Wed Jul 20 15:45:55 EDT 2011


I second this.  If your machine has "stability issues" then all the
invariants and guarantess that ZFS provides are voided, because the data
written by ZFS on disk is just not reliable anymore.  Say, an uberblock
got corrupted or never got written despite the hardware claiming it
did... ZFS can't do anything about it.

FIX YOUR BOX.  FIRST.

On Tue, 2011-07-19 at 02:42 +0000, Steve Costaras wrote:
> 
> 
> 
> As for your issue specifically:
> - First thing fix your PC for stability.
> - Second make sure that you can see all drives that you had that
> formed the particular pool (/proc/partitions  ; /dev/sd{x} ;
> or /sys/block/{x}).
> - Third, if possible would be to create 1:1 bit copies of all your ZFS
> drives so you can always get back to your starting point.:
>       dd if=/dev/{disk} of=/mnt/{disk}.img bs=512 conv=noerror,sync
>       (bs=4096 if you have ashift=12/newer AFT drive).
> 
> Once that's done you can try using the solaris 11 live CD to see if it
> can import your data.    If that fails it's more of a dive down
> forensic lane.  Unforutnatly, forensic tools for ZFS are very lacking,
> best would be to probably do this off the solaris live CD as most
> tools are writen w/ solaris in mind for ZFS.   Here's a link for some
> starting info:  http://hub.opensolaris.org/bin/view/Project
> +forensics/ZFS-Forensics 
> 
> 
> Though you don't want to here this, this happens enough that it needs
> to be repeated:  No file system or single solution can protect
> critical data, always have backups. 
> 
> 
>         -----Original Message-----
>         From: UbuntuNewbie [mailto:chessdoter at googlemail.com]
>         Sent: Monday, July 18, 2011 09:16 PM
>         To: 'zfs-discuss'
>         Subject: [zfs-discuss] !HELP! Lost my pool?!?
>         
>         I was/am running the latest version from PPA. Due to non-zfs
>         related issues, i am dealing with instabilities, seeing the PC
>         rebooting by itself from time to time. As this just happened,
>         this time my pool doesnt import anymore, stating, it would be
>         faultet (corrupted data). Solaris docs state, the pool needs
>         to be destroyed and recreated. That would be the worst data
>         loss since years (on the "safest FS" i ever used), because i
>         did not yet backup the pool (i only have a backup of the OS
>         partition). About 500 GB worth of data is in the pool, some of
>         it cannot be recreated if you cannot guide me to some saving
>         measures or show me how to import what is left and healthy. Of
>         course, i could make a low- level copy of the partition, since
>         the pool only uses a slice. But i cannot read from that data.
>         Quite a shocker!




More information about the zfs-discuss mailing list