Tweaking/Optimizing MD RAID: 195MB/s write, 181MB/s read (so far)

From
Justin Piszcz <hidden>
Date
2007-01-11 23:05:36
Also in
linux-xfs
With 4 Raptor 150s & XFS (default XFS options):

# Stripe tests:
echo 8192 > /sys/block/md3/md/stripe_cache_size

# DD TESTS [WRITE]

DEFAULT:
$ dd if=/dev/zero of=10gb.no.optimizations.out bs=1M count=10240
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB) copied, 96.6988 seconds, 111 MB/s

8192 STRIPE CACHE
$ dd if=/dev/zero of=10gb.8192k.stripe.out bs=1M count=10240
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB) copied, 55.0628 seconds, 195 MB/s
$

16384 STRIPE CACHE
$ dd if=/dev/zero of=10gb.16384k.stripe.out bs=1M count=10240
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB) copied, 56.2793 seconds, 191 MB/s

# DD TESTS [READ]

DEFAULT:
$ dd if=10gb.16384k.stripe.out of=/dev/null bs=1M
298+0 records in
297+0 records out
311427072 bytes (311 MB) copied, 3.5453 seconds, 87.8 MB/s

2048K READ AHEAD
$ dd if=10gb.16384k.stripe.out of=/dev/null bs=1M
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB) copied, 85.4632 seconds, 126 MB/s

8192K READ AHEAD
$ dd if=10gb.16384k.stripe.out of=/dev/null bs=1M
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB) copied, 64.9454 seconds, 165 MB/s

16384K READ AHEAD
$ dd if=10gb.16384k.stripe.out of=/dev/null bs=1M
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB) copied, 59.3119 seconds, 181 MB/s

Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help