btrfs RAID1 benchmark on Dell E4200 with 128GB SSD

Having previously posted some nice XFS Bonnie++ numbers on the work Dell E4200 I use I thought I’d redo these after having migrated to a RAID-1 configuration of the experimental btrfs filesystem. As SSD’s are not necessarily as reliable as spinning disk yet for data integrity I wanted a system that could spot this and correct for it, so I created two equal size partitions on the SSD and created the filesystem with mkfs.btrfs -d raid1 /dev/sda10 /dev/sda11 before mounting it with the ssd mount option.

Version 1.03c       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
sys26            2G           27458  23 21782  19           108478  39  3079  21

                    ------Sequential Create------ --------Random Create--------
                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16  7542  99 +++++ +++  5818  99  7232  99 +++++ +++  5638 100
sys26,2G,,,27458,23,21782,19,,,108478,39,3078.8,21,16,7542,99,+++++,+++,5818,99,
7232,99,+++++,+++,5638,100

real    3m30.808s
user    0m0.404s
sys     0m55.599s

Whilst the raw numbers are nowhere near as good you have to remember that this is doing checksumming of all the data and mirroring it across the two partitions I created for it and, if it finds a problem with the data, will try and recover using the data on the other partition. It’s also still in development!

3 thoughts on “btrfs RAID1 benchmark on Dell E4200 with 128GB SSD

  1. >Whilst the raw numbers are nowhere near as good you have to remember that this is doing checksumming of all the data and mirroring it across the two partitions I created for it and, if it finds a problem with the data, will try and recover using the data on the other partition. It’s also still in development!

    What, like ZFS? 😛

  2. Chris Mason pointed out on the mailing list that RAIDing multiple partitions will often not physically separate data being written to a SSD, and thus may not provide redundancy. Would it be too much trouble to ask for you to re-run without RAID1? I’d also like to see XFS added for good measure.

Comments are closed.