| To: | "Thomas von Steiger" <thomas.vonsteiger@xxxxxxxxxx> |
|---|---|
| Subject: | Re: lost+found |
| From: | Steve Lord <lord@xxxxxxx> |
| Date: | Thu, 21 Jun 2001 18:02:08 -0500 |
| Cc: | "Deti Fliegl" <deti@xxxxxxxxx>, "Joshua Baker-LePain" <jlb17@xxxxxxxx>, linux-xfs@xxxxxxxxxxx |
| Comments: | In-reply-to "Thomas von Steiger" <thomas.vonsteiger@bluewin.ch> message dated "Fri, 22 Jun 2001 00:56:43 +0200." |
| References: | <Pine.LNX.4.33.0106161320550.27379-100000@chaos.egr.duke.edu> <3B2B9A04.12BDDC7A@fliegl.de> <01f901c0faa5$71404880$6402a8c0@rubistar> |
| Sender: | owner-linux-xfs@xxxxxxxxxxx |
> Thanks, > > If a xfs Filesystem is 100% full, is it posible for xfs_repair to build a > temp lost+found ? > Repair is wunderfull, if i have a power interrupt...!! > > regards, > thomas > Well, since lost+found is relinking existing files from elsewhere which have become disconnected it potentially does not consume any space at all, presuming that: o there is room in the root directory for one extra entry without adding a new disk block. o there is a spare inode somewhere - remember these are allocated in chunks and never freed, so chances are good. o All the lost+found entries fit in the directory inode. But even if all of these fail, 4 disk blocks is probably all it takes at maximum, maybe more if there are lots of entries to recover. So it depends on your definition of full. Steve |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: lost+found, Thomas von Steiger |
|---|---|
| Next by Date: | Re: NFS bugs w/ SGI's, Steve Lord |
| Previous by Thread: | Re: lost+found, Thomas von Steiger |
| Next by Thread: | Newer patches?, Patrik Schindler |
| Indexes: | [Date] [Thread] [Top] [All Lists] |