kern/137797: gmirror split does not improve performance
don_oles at able.com.ua
Sat Aug 15 14:20:07 UTC 2009
>Synopsis: gmirror split does not improve performance
>Arrival-Date: Sat Aug 15 14:20:06 UTC 2009
>Originator: Oles Hnatkevych
>Release: FreeBSD 7.2
Mirror is created with gmirror on ad8 and ad10
ad8: 476940MB <Seagate ST3500630NS 3.AEK> at ata4-master SATA150
ad10: 476940MB <Seagate ST3500630NS 3.AEK> at ata5-master SATA150
Geom name: ar1
mirror/ar1 COMPLETE ad8
The split algorithm does not improve any performance. Even simple dd with block size 1M has the same speed as a single drive, while two dd's on singe drive run simultaneously proove that SATA throughput is not a bottleneck.
gmirror label ar1 ad8 ad10
dd if=/dev/ad8 of=/dev/null bs=1M count=500
dd if=/dev/mirror/ar1 of=/dev/null bs=1M count=500
both give the same speed (around 80MB/s for Seagate 500GB SATA drives)
dd if=/dev/ad8 of=/dev/null bs=1M count=500 & dd if=/dev/ad10 of=/dev/null bs=1M count=500
show that in parallel they do provide data faster.
BTW, the performance of a RAID1 created with the atacontrol is a bit better then gmirror.
More information about the freebsd-bugs