[rescue] Sun DiskSuite RAID problems

Patrick Giagnocavo xemacs5 at gmail.com
Sat Apr 27 13:18:48 CDT 2019


I would strongly suggest using ddrescue to image the "Maintenance" state
drive.

I had a big problem once, RAID1 pair that died on me.  Not only were the
drive the same make and model but were consecutive serial numbers!

I used ddrescue to move the drives to another system and only lost 32KB of
data across 200+GB of data. Never figured out what it was, maybe part of an
image, a man page, etc. However everything came up and worked after that.

Cheers

Patrick

On Sat, Apr 27, 2019 at 9:50 AM Carl R. Friend <crfriend at rcn.com> wrote:

>     Good morning Rescuers,
>
>     I figured I'd give this one more shot before abandoning all hope on
> the matter.
>
>     I've got a dead DiskSuite array created on Solaris 8 that has one
> verifiably dead drive (that ain't comin' back) and one that's in the
> "Maintenance" state but which reads OK.  On this array is/was a large
> archive of GOES-13 satellite images going back to 2004 running until
> GOES-13 was decommissioned (as well as a bunch of other interesting
> stuff).
>
>     "Is there a way to force the "Maintenance" drive out of that state
> and into a "good" state without invoking the RAID-rebuild routines?"
>
>     The good news is that I was able to get what look like good images
> off all of the other disks so another avenue might be to write something
> to "rebuild" the contents of the dead drive and copy that image to a
> different disk and install that one in the original's place.  This
> begs a second question:
>
>     "What algorithm did Sun use for the RAID-rebuild tactic?"  If it's
> something simple, I might be able to do something "creative" to make
> an image of the failed disk.  If it's something ridiculously complex,
> then I suppose I should give up, lick my wounds, and move on.  All the
> disk geometries are known and documented, the RAID configuration is
> known and documented, the location of the metadevice-databases is known
> and documented.  This should be possible.
>
>     Ideas?  Abuse?  Suggestions?
>
>     Some level of corruption is definitely anticipated, but the key
> parts are likely OK so long as the directory-structures are intact.
>
>     Cheers!
> --
> +------------------------------------------------+---------------------+
> | Carl Richard Friend (UNIX Sysadmin)            | Boylston            |
> | Minicomputer Collector / Enthusiast            | Massachusetts, USA  |
> | mailto:crfriend at rcn.com                        +---------------------+
> | http://users.rcn.com/crfriend/museum           | ICBM: 42:20N 71:43W |
> +------------------------------------------------+---------------------+
> _______________________________________________
> rescue list - http://www.sunhelp.org/mailman/listinfo/rescue


More information about the rescue mailing list