journaled quota file question

journaled quota file question

[Dmitry Monakhov]

While looking to ext3/4 quota code I'm wondering
Why do we have to place journaled quota file on fs root?
In some situations it may be useful to place it deeper for example:
/root/dir/quota_file
The only reason what comes in to my mind is that some one
may rename parent folder rename("/root/dir", "/root/dir2")
Is this the only reason?

Re: journaled quota file question

[Dmitry Monakhov]

Dmitry Monakhov <dmonakhov@openvz.org> writes:

> While looking to ext3/4 quota code I'm wondering
> Why do we have to place journaled quota file on fs root?
> In some situations it may be useful to place it deeper for example:
> /root/dir/quota_file
> The only reason what comes in to my mind is that some one
> may rename parent folder rename("/root/dir", "/root/dir2")
> Is this the only reason?
add an author in cc:
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: journaled quota file question

[Jan Kara]

  Hi,

On Sat 31-10-09 21:59:44, Dmitry Monakhov wrote:
> Dmitry Monakhov <dmonakhov@openvz.org> writes:
> 
> > While looking to ext3/4 quota code I'm wondering
> > Why do we have to place journaled quota file on fs root?
> > In some situations it may be useful to place it deeper for example:
> > /root/dir/quota_file
> > The only reason what comes in to my mind is that some one
> > may rename parent folder rename("/root/dir", "/root/dir2")
> > Is this the only reason?
  Well, the reason is that we have to read & write the files during orphan
recovery on mount. A that time the filesystem is not fully set up so you
cannot use VFS directory traversal code and I didn't want to implement
ext[34] specific one. Also as you write above, if you have quota files in
some subdirectory, there are more possibilities for failure.
  Admittedly, I don't see a big point in having quota files in some other
directory. In fact, they should rather be system files not visible in
any directory but that requires rather non-trivial changes to repquota and
moving quotacheck functionality to fsck and I never got to doing that.

> add an author in cc:
  Thanks.

									Honza
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR