[NCLUG] file system damage from dd with of=<a drive in my file system>

[Stephen Warren]

Before you do anything, you should take an image of both drives onto a
third unrelated drive so nothing gets worse.

To find out exactly which physical volumes contain (various parts of)
your logical volumes, you can do a couple of things:

a) Run the command: "lvdisplay --maps".

This will list all the LVs in your system, and for each, a list of
segments showing where those segments are stored.

b) Look at the file "/etc/lvm/backup/${vg_name}"

(replace "${vg_name}" with the name of your VG)

Again, the "logical_volumes" section contains an entry for each LV, and
within each of those, there is a list of "segment" sections that
indicates where the data is stored.


You might be lucky; even though your VG includes PVs from both physical
disks, unless your LV is large enough not to fit into the first PV, it's
possible the second PV isn't used yet.

Of course, hopefully you have backups just in case.

On 11/22/2015 09:32 AM, Ryan Clemens wrote:
> You've almost certainly overwritten the partition table on sdb. As long as
> your system boots and starts up you probably had no real data on it before,
> so you're fine to keep using your current red hat install.
> 
> I'd recommend using "parted" (or your favorite red hat dish utility) to
> repartition sdb however you wish and then modify your etc/fstab file (or
> some red hat dish utility) to mount the partition at boot to be able to use
> it again.
> On Nov 22, 2015 09:17, "Ray Frush" <phred at rflm.net> wrote:
> 
>> Well, that was unexpected.   A default Redhat install will use LVM to
>> allocate space on /dev/sda, so that’s expected, but it’s been a while since
>> I’ve installed on a system with two disks.  I didn’t expect it to use
>> /dev/sdb as well.  (Usually, I’ll add in a second disk after the base
>> install)   Did the disk layout get customized during install?
>>
>>
>> In any case, your Volume Group is made up of physical volumes from both
>> /dev/sda2 and /dev/sdb1, and then I’m guessing the Logical Volumes were
>> allocated out of that combination of space.   Since you’ve overwritten
>> /dev/sdb (and /dev/sdb1) with the contents of a large ISO file,  I wouldn’t
>> expect any data on /dev/sdb1 to be recoverable.
>>
>> —
>> Ray Frush
>>
>>
>>> On 22 Nov 2015, at 8:50 AM, Vincent Randal <vtrandal at gmail.com> wrote:
>>>
>>> Thanks Ray, you given me a good idea where to look for the problem.
>>>
>>> sudo vgdisplay -v | tail -n 11
>>>    Using volume group(s) on command line. # what?
>>>  --- Physical volumes ---
>>>  PV Name               /dev/sda2
>>>  PV UUID               9KDSFf-xdes-deo2-18nd-Z3GW-VLlq-JMbmiL
>>>  PV Status             allocatable
>>>  Total PE / Free PE    59257 / 0
>>>
>>>  PV Name               /dev/sdb1
>>>  PV UUID               Ceh1Cc-aGH7-voMx-rbpR-fXAj-MvOr-sWEqwn
>>>  PV Status             allocatable
>>>  Total PE / Free PE    121990 / 0
>>>
>>> sudo vgck
>>>  Couldn't find device with uuid Ceh1Cc-aGH7-voMx-rbpR-fXAj-MvOr-sWEqwn.
>>>  The volume group is missing 1 physical volumes.
>>>
>>> sudo pvck /dev/sdb1
>>>  Could not find LVM label on /dev/sdb1
>>>
>>> Maybe I can fix the problem from steps described here?
>>>
>> https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/6/html/Logical_Volume_Manager_Administration/mdatarecover.html
>>> _______________________________________________
>>> NCLUG mailing list       NCLUG at lists.nclug.org
>>>
>>> To unsubscribe, subscribe, or modify
>>> your settings, go to:
>>> http://lists.nclug.org/mailman/listinfo/nclug
>>
>>
>> _______________________________________________
>> NCLUG mailing list       NCLUG at lists.nclug.org
>>
>> To unsubscribe, subscribe, or modify
>> your settings, go to:
>> http://lists.nclug.org/mailman/listinfo/nclug
>>
> _______________________________________________
> NCLUG mailing list       NCLUG at lists.nclug.org
> 
> To unsubscribe, subscribe, or modify
> your settings, go to:
> http://lists.nclug.org/mailman/listinfo/nclug
>