On Sat, May 30, 2009 at 02:40:34PM +0100, Neil Bothwick wrote: > On Sat, 30 May 2009 12:06:04 +0200, Florian Philipp wrote: > > > Delaying commits with ext4 and/or laptop-mode will reduce the wear-down > > of your SSD but it might as well freeze your system when the actual > > commit takes place because these things tend to have a terribly low > > write performance. > > That may explain the pauses I get from time to time. Maybe shortening the > commit period will help.
Couple of points regarding the pauses, SSDs, schedulers and ext3/ext4: * try ext4 with its delayed allocation. It should help with pauses * ext3 with data=writeback should help. Some security implications with data=writeback tho. So be careful if it is not a single user machine. * Deadline scheduler has more throughput than CFQ or anticipatory but it is totally unusable under load * A lot of patches to ext3 and ext4 for a/m pauses and SSDs. Some made it to kernel 2.6.30 I believe. * Try CFQ and NOOP as schedulers for SSDs for now. After the above patches, CFQ should be the better choice. Basically, a lot of changes to ext3/ext4 and schedulers at the moment. I would wait for at least kernel 2.6.31 before trying alternatives and making decisions. > Or I could try btrfs, which has an ssd mount option. Ugh. Even on-disk format is not finalized yet. -- Eray
