| To: | pille <pille+xfs+mailinglist+sgi@xxxxxxxxxxxx> |
|---|---|
| Subject: | Re: du vs. ls |
| From: | Matthias Schniedermeyer <ms@xxxxxxx> |
| Date: | Fri, 4 Jan 2013 17:14:49 +0100 |
| Cc: | xfs@xxxxxxxxxxx |
| In-reply-to: | <50E6F6D4.5090407@xxxxxxxxxxxx> |
| References: | <50E6F6D4.5090407@xxxxxxxxxxxx> |
| User-agent: | Mutt/1.5.21 (2010-09-15) |
On 04.01.2013 16:35, pille wrote: > hi, > > # du -hs file copy > 128M file !! > 100M copy I think you are bitten by "speculative preallocation". When XFS "thinks" you will extent the file in the future it speculativly allocates space. This prevents fragmentation. This fixes itself over time. Either do enough IO that the cache of the copy gets reused or umount or "echo 3 > /proc/sys/vm/drop_caches" (The last one drops the whole cache (except dirty pages) of Linux!) If it is a problem, the behaviour can be disabled by using the mount-option "allocsize", for example: "allocsize=4k" -- Matthias |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: du vs. ls, Ben Myers |
|---|---|
| Next by Date: | Re: [PATCH] xfs: fix _xfs_buf_find oops on blocks beyond the filesystem end, Dave Chinner |
| Previous by Thread: | Re: du vs. ls, Ben Myers |
| Next by Thread: | Re: du vs. ls, pille |
| Indexes: | [Date] [Thread] [Top] [All Lists] |