zfs benchmarks and 7 disk raidz oddity

[Andrew Snow-2]

Hi all,

I am running 8-CURRENT with ZFS patches on a 3ghz Core2Duo with Intel
P45 chipset.  I took some benchmarks of ZFS on six old SATA disks and
one PATA, onboard controllers only, and I expect you are interested to
see the results.  The disks range in size from 200gb to 320gb.

I tried several configurations with quick and dirty testing. Listed
below is the sequential megabytes/sec rating as measured by dd bs=1m for
a 10GB file.  All ZFS settings were left at their defaults.


Conf                    Write   Read (MB/s)
------------------------------------
7 disk RAIDZ2           220     305
7 disk RAIDZ1            84     361
7 disk striped          318     409
7 disk stripe copies=2  140     164

6 disk RAIDZ2           173     260
6 disk RAIDZ1           238     307
6 disk striped          280     346
6 disk 2xRAIDZ1 striped 188     251
6 disk striped mirrors  164     323
6 disk stripe copies=2  151     179


A few notes:

1. using copies=2 is a nice way to be able to get RAID1-like mirroring
reliability but on an odd number of disks.  However you take a
noticeable performance penalty: Write speed is fine but read speed is
almost half of what RAID0+1 achieved.

2. RAIDZ1 and RAIDZ2 are fast and efficient.  However in total it caused
the system to use almost all of one CPU core during writing.

3. There seemed to be a bug with 7 disks and RAIDZ1 - the write
performance was terrible!  When I ran "gstat" I noticed it was spending
much time writing to only two disks most of the time, which became a
serious bottleneck - the worst write score of the lot.  Read was fine.
Perhaps the algorithm isn't optimised for choosing parity locations out
of exactly 7 disks?



- Andrew

_______________________________________________
freebsd-fs@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-unsubscribe@..."

[Ruslan Kovtun]

Hi.

Thank you very much, very useful information.
It is interesting  when we can use this on 7.x?


В сообщении от Суббота 20 сентября 2008 Andrew Snow написал(a):

> Hi all,
>
> I am running 8-CURRENT with ZFS patches on a 3ghz Core2Duo with Intel
> P45 chipset.  I took some benchmarks of ZFS on six old SATA disks and
> one PATA, onboard controllers only, and I expect you are interested to
> see the results.  The disks range in size from 200gb to 320gb.
>
> I tried several configurations with quick and dirty testing. Listed
> below is the sequential megabytes/sec rating as measured by dd bs=1m for
> a 10GB file.  All ZFS settings were left at their defaults.
>
>
> Conf                    Write   Read (MB/s)
> ------------------------------------
> 7 disk RAIDZ2           220     305
> 7 disk RAIDZ1            84     361
> 7 disk striped          318     409
> 7 disk stripe copies=2  140     164
>
> 6 disk RAIDZ2           173     260
> 6 disk RAIDZ1           238     307
> 6 disk striped          280     346
> 6 disk 2xRAIDZ1 striped 188     251
> 6 disk striped mirrors  164     323
> 6 disk stripe copies=2  151     179
>
>
> A few notes:
>
> 1. using copies=2 is a nice way to be able to get RAID1-like mirroring
> reliability but on an odd number of disks.  However you take a
> noticeable performance penalty: Write speed is fine but read speed is
> almost half of what RAID0+1 achieved.
>
> 2. RAIDZ1 and RAIDZ2 are fast and efficient.  However in total it caused
> the system to use almost all of one CPU core during writing.
>
> 3. There seemed to be a bug with 7 disks and RAIDZ1 - the write
> performance was terrible!  When I ran "gstat" I noticed it was spending
> much time writing to only two disks most of the time, which became a
> serious bottleneck - the worst write score of the lot.  Read was fine.
> Perhaps the algorithm isn't optimised for choosing parity locations out
> of exactly 7 disks?
>
>
>
> - Andrew
>
> _______________________________________________
> freebsd-fs@... mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-fs
> To unsubscribe, send any mail to "freebsd-fs-unsubscribe@..."

--
________________
Ruslan Kovtun
mailto: yalur@...
mob: [phone]
ICQ: 277696182
_______________________________________________
freebsd-fs@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-fs
To unsubscribe, send any mail to "freebsd-fs-unsubscribe@..."