|To:||Dave Chinner <david@xxxxxxxxxxxxx>|
|Subject:||Re: TRIM on XFS|
|From:||Jeffrey Hundstad <jeffrey.hundstad@xxxxxxxx>|
|Date:||Tue, 11 Oct 2011 11:06:24 -0500|
|References:||<201110071352.20135@xxxxxx> <20111007233450.GK3159@dastard> <201110100829.45823@xxxxxx> <CAGpXXZJs8kjZaSp248rA+-d=gpwWe=oK0xtUUirv+zgE3Vq3kA@xxxxxxxxxxxxxx> <4E936BFB.5060907@xxxxxxxx> <20111011020552.GY3159@dastard>|
|User-agent:||Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:220.127.116.11) Gecko/20111005 Icedove/3.1.15|
Dave,Thank you. I suspected this may be the case. It is awfully nice to have a definitive answer though. It is especially nice to have a way to determine if it is possible. Thank you!
I've created a an XFS Wiki page for this info: http://xfs.org/index.php/FITRIM/discard -- Jeffrey Hundstad On 10/10/2011 09:05 PM, Dave Chinner wrote:
On Mon, Oct 10, 2011 at 05:04:43PM -0500, Jeffrey Hundstad wrote:On 10/10/2011 03:28 PM, Greg Freemyer wrote:fitrim is the batch mode. To enable realtime discard, add --discard to your mount options. Documented athttp://lxr.linux.no/#linux+v3.0.4/Documentation/filesystems/xfs.txt To invoke batch mode, simply use fstrim from user space.Hello, I'm also trying to use FITRIM on a simple XFS partition on a spinning magnetic hard drive (remember those). I haven't had much luck. I'm doing this just to get used to the commands when using this against a thinly provisioned disk. I have no idea if this should work in this configuration. On a Debian system with Linux 3.0.0-2-amd64 when I issue: # fstrim / I receive: fstrim: /: FITRIM ioctl failed: Operation not supported With an strace this looks like: open("/", O_RDONLY) = 3 ioctl(3, 0xc0185879, 0x7fff4ea851b0) = -1 EOPNOTSUPP (Operation not supported)It means the block device underneath the filesystem doesn't support the TRIM operation. Check /sys/block/<dev>/queue/discard_max_bytes - if the value is zero then your device doesn't support discard operations. Cheers, Dave.
|<Prev in Thread]||Current Thread||[Next in Thread>|
|Previous by Date:||[PATCH 1/4] xfs: do not update xa_last_pushed_lsn for locked items, Christoph Hellwig|
|Next by Date:||Re: 18.104.22.168 kernel bug in XFS or megaraid driver with heavy I/O load, Jesper Krogh|
|Previous by Thread:||Re: TRIM on XFS, Dave Chinner|
|Next by Thread:||Re: TRIM on XFS, Dave Chinner|
|Indexes:||[Date] [Thread] [Top] [All Lists]|