Scrub Quesiton

[Thomas Blomberg Hansen]

My new(onboard Supermicro Q67) hardware is a bit faster than the old(Adaptec 1430SA + onboard Intel H67).

I did the same dd (did them twice) on my new system for you to compare Raidflex:

Write:
[root@freenas] /mnt/data# dd if=/dev/zero of=/mnt/data/test.dat bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 266.016165 secs (157671020 bytes/sec)
Read:
[root@freenas] /mnt/data# dd if=/mnt/data/test.dat of=/dev/null bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 286.713548 secs (146289006 bytes/sec)

Write:
[root@freenas] /mnt/data# dd if=/dev/zero of=/mnt/data/test.dat bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 284.435178 secs (147460804 bytes/sec)
Read:
[root@freenas] /mnt/data# dd if=/mnt/data/test.dat of=/dev/null bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 301.711236 secs (139017163 bytes/sec)

 

[b1ghen]

Just as a comparison I did the exact same dd tests on my RAIDZ2 pool:

Write:
[root@freenas] ~# dd if=/dev/zero of=/mnt/tank/test.dat bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 180.038561 secs (232966981 bytes/sec)

Read:
[root@freenas] ~# dd if=/mnt/tank/test.dat of=/dev/null bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 137.185264 secs (305740127 bytes/sec)

Are your pools seeing a lot of other traffic/usage while doing the scrub/benchmarks?
My system is usually idling or at most straming a single 720p video over SMB.

 

[raidflex]

Just as a comparison I did the exact same dd tests on my RAIDZ2 pool:

Write:
[root@freenas] ~# dd if=/dev/zero of=/mnt/tank/test.dat bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 180.038561 secs (232966981 bytes/sec)

Read:
[root@freenas] ~# dd if=/mnt/tank/test.dat of=/dev/null bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 137.185264 secs (305740127 bytes/sec)

Are your pools seeing a lot of other traffic/usage while doing the scrub/benchmarks?
My system is usually idling or at most straming a single 720p video over SMB.

I had to leave the system idle while running the scrub, because anytime I would access the pool it was be extremely slow just browsing files.

 

[raidflex]

I ended up purchasing a m1015 controller card on eBay for $58 brand new.

What do I need to do with my zfs pool when moving to a new controller?

 

[peterh]

You need to "export" the disks ( man zpool) and there is a GUI function to do this, BUT LOOK OUT, there is
also the choice to erase the disks.
Then you need to import them.
But as you are not moving between systems, chances are that you just have to shutdown-change hw-reboot.
Any takers with experience of this subject ?

 

[Thomas Blomberg Hansen]

I just moved my pool to my new motherboard = new controller :)

Did a scrub ;P
Exported
Auto imported the pool when I was sure the system was okay.

Easy :)

Ps. the scrub was only to sure I did all I could.
You just made yours and with no errors. So I don't think its nessary.

I waiting to see your result. My change did almost nothing.

@b1ghen: My system aren't doing anything, maybe Sickbeard downloaded a .torrent 40-50 KB
Transmission was shutdown and I didn't saw any Tvshow or movie.

I don't need more speed for the daily routine. But I want my system to perform what it should able to :)

PPS: the Antec P280 is a really nice tower lots of room, nice cabling and the hdd trays are perfect :)
The Supermicro Q67 mATX motherboard is perfect :)

 

[raidflex]

Yeah same here I may not always need the speed but I want it there when I do.

I have an older Antec 4u rack mount case with 8 hot swap bays I put together years ago. Antec makes some nice cases.

I will report in when I get the new controller up and running and do some tests.

 

[raidflex]

Well I installed the new M1015 controller and flashed to IT firmware.

My READ/WRITE speeds have doubled!


Write

root@fileserver] ~# dd if=/dev/zero of=/mnt/Data/test.dat bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 179.765640 secs (233320672 bytes/sec)


Read

[root@fileserver] ~# dd if=/mnt/Data/test.dat of=/dev/null bs=8M count=5000
5000+0 records in
5000+0 records out
41943040000 bytes transferred in 61.367187 secs (683476662 bytes/sec)

I will run a scrub sometime in the next couple days and report back in.

 

[raidflex]

What a difference in the scrub time! Its 1/10th the time!


scrub: scrub completed after 3h58m with 0 errors on Mon Mar 26 00:53:51 2012
config:

NAME STATE READ WRITE CKSUM
Data ONLINE 0 0 0
raidz2 ONLINE 0 0 0
gptid/e902c40d-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e91172b0-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e91f0e06-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e92cba1e-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e93a03e7-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e94867a0-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e956b27d-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0
gptid/e964d6cf-3326-11e1-acd0-e06995ebe0de ONLINE 0 0 0

errors: No known data errors