Yes, that would be cool. Brock
-----Original Message----- From: Sage Weil [mailto:[EMAIL PROTECTED] Sent: Thursday, August 14, 2008 7:10 PM To: Erwin, Brock A Cc: ceph-devel@lists.sourceforge.net Subject: Re: [ceph-devel] Striping strategy for ceph Hi Brock- On Thu, 14 Aug 2008, Erwin, Brock A wrote: > I understand that ceph defaults to stripe in 1 Meg chunks over OSDs, > however, is it possible to force ceph to store a particular file on one > and only one OSD? I am interested in doing some work with Active > Storage, which seeks to do processing on the data on the OSD itself and > not on the client that mounts the filesystem. In traditional parallel > file systems, OSDs are usually left with a lot of processing power > unused. If the files are striped, then active storage becomes extremely > complex. So in conclusion, it would be nice if ceph were customizable > in that it would allow you to turn off striping for any files that you > specify. All of the data structures and infrastructure is in place for objects to be preferentially allocated to a specific OSD on a per-file basis. The only missing piece is the interface to let users set it.. an ioctl() to specify the layout (including things like stripe size) after the file is created (but before data is written). If you'd like to play with this, I'll move that up on my todo list! sage ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Ceph-devel mailing list Ceph-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ceph-devel
