[Top] [All Lists]

Re: xfs_repair segfaults

To: Ole Tange <tange@xxxxxxxxxx>
Subject: Re: xfs_repair segfaults
From: Eric Sandeen <sandeen@xxxxxxxxxxx>
Date: Mon, 04 Mar 2013 09:20:04 -0600
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <CANU9nTkekrw2mdptFdK2RVUFXD=evt0sTZhB9VL+PKY8DTj11A@xxxxxxxxxxxxxx>
References: <CANU9nTnvJS50vdQv2K0gKHZPvzzH5EY1qpizJNsqUobrr2juDA@xxxxxxxxxxxxxx> <512FA67D.2090708@xxxxxxxxxxx> <CANU9nTm3vR3Z5JwRwou4vnNejstAkAsnBb2ivwx-KhnHHbWpoQ@xxxxxxxxxxxxxx> <5130DB54.9030503@xxxxxxxxxxx> <CANU9nTkekrw2mdptFdK2RVUFXD=evt0sTZhB9VL+PKY8DTj11A@xxxxxxxxxxxxxx>
User-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130216 Thunderbird/17.0.3
On 3/4/13 3:00 AM, Ole Tange wrote:
> On Fri, Mar 1, 2013 at 5:46 PM, Eric Sandeen <sandeen@xxxxxxxxxxx> wrote:
>> On 3/1/13 3:37 AM, Ole Tange wrote:
>>> Please consider providing an example in the man page for xfs_metadump e.g:
>>>   xfs_metadump.sh -g /dev/sda2 meta.dump
>> From the manpage,
>>        xfs_metadump [ -efgow ] [ -l logdev ] source target
>> The source argument must be the pathname of
>> the device or file containing the XFS filesystem
>> and
>> the target argument specifies the destination file name.
>> is not enough?
> I have never run xfs_metadump before and I am in a state of worry that
> my filesystem is toast. I would therefore like to be re-assured that
> what I am doing is correct. I did that by reading and re-reading the
> manual to make sure I had understood it correctly. By providing me
> with an example of the right way to do it in the man page, I will feel
> more confident that what I am about to do it correct and I could
> probably save time by not having to re-read the manual.
> So I am not saying the information is not there, what I am saying is
> that you in a simple way could make it easier to grasp the
> information.

Fair enough, maybe a concrete example is warranted.

I suggested the meatadump for 2 reasons:

1) to get an image we could look at, to analyze the reason for the segfault, and
2) so you could run a "real" non-"n" xfs_repair on a metadata image as a more 
realistic dry run

xfs_metadump only *reads* your filesystem, so there is nothing dangerous.

But I understand your paranoia and worry.  :)


> /Ole

<Prev in Thread] Current Thread [Next in Thread>