On 12 November 2010 22:14, Peter Kerwien <peter%kerwien.homeip.net@localhost>
wrote:
> Any comments on the following:
>
> Same disks, HW but different OS and filesystem: Linux Mint 10 (amd64)
> with mdadm created raid-5 and ext4. Chunk size = 32KB, block size = 4KB,
> stride size = 8 blocks = 32KB.
>
> # dd if=/mnt/vol1/pub/foo.mkv of=/dev/null bs=100M count=100
> 100+0 records in
> 100+0 records out
> 10485760000 bytes (10 GB) copied, 100.641 s, 104 MB/s
>
> # dd if=/dev/zero of=/mnt/vol1/pub/zero bs=100M count=100
> 100+0 records in
> 100+0 records out
> 10485760000 bytes (10 GB) copied, 102.646 s, 102 MB/s
>
> Note the *slightly* higher write speed. Comments?
>
I've noticed that with the same RAID configuration as yourself, I
started the wedge at 128 sectors, not 64 (ie a whole stripe of data
from the start), I remember having terrible performance until I got
this bit correct.
I have 3x2T HGST disks, set up like the following:-
(ian:~)$ sudo raidctl -G raid0
# raidctl config file for /dev/rraid0d
START array
# numRow numCol numSpare
1 3 0
START disks
/dev/wd4a
/dev/wd2a
/dev/wd3a
START layout
# sectPerSU SUsPerParityUnit SUsPerReconUnit RAID_level_5
64 1 1 5
START queue
fifo 100
(ian:~)$ sudo gpt show raid0
start size index contents
0 1 PMBR
1 1 Pri GPT header
2 32 Pri GPT table
34 94
128 7814057951 1 GPT part - NetBSD UFS/UFS2
7814058079 32 Sec GPT table
7814058111 1 Sec GPT header
(ian:~)$ sudo dkctl raid0 listwedges
/dev/rraid0d: 1 wedge:
dk0: 16cdb564-4432-11df-8352-0050430072f6, 7814057951 blocks at 128, type: ffs
(ian:~)$
This gives the following performance:
(ian:~)$ cd /usr3/share/
(ian:/usr3/share)$ dd if=/dev/zero of=zero bs=100m count=1k
1024+0 records in
1024+0 records out
107374182400 bytes transferred in 626.305 secs (171440723 bytes/sec)
(ian:/usr3/share)$ dd if=zero of=/dev/null bs=100m count=1k
1024+0 records in
1024+0 records out
107374182400 bytes transferred in 458.915 secs (233974009 bytes/sec)
(ian:/usr3/share)$
(Quad core Phenom on a 780G motherboard.)
Cheers,
Ian