RE: Orion, NetApp, and iops
Date: Tue, 4 May 2010 12:34:00 -0400
Message-ID: <56211FD5795F8346A0719FEBC0DB0675067CDED8_at_mds3aex08.USIEXCHANGE.COM>
Sounds to me, that if your calcs are correct on iops per disk, then the speed different is mostly due to caching. You need to figure out how much cache is on the netapp side. If you are trying to bypass the cache, then make your sample files far bigger than cache and the Orion runs longer.
-----Original Message-----
From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Henry Poras
Sent: Tuesday, May 04, 2010 10:19 AM
To: Jakub Wartak
Cc: oracle-l_at_freelists.org
Subject: Re: Orion, NetApp, and iops
Sure.
The OS is Linux (2.6.18) with FC. Orion is bundled with an 11.2 install.
The NetApp filer is a 3050.
I'm not sure about the detailed NetApp settings and would have to find where to look. But even if various optimizations are turned on, I don't understandthese results. It might be read ahead, but if Orion is really random reads across the disk I wouldn't have expected this to show up.
I will try the "stats show disk" command as soon as I can. That should help.
I also believe that RAID-DP shouldn't lead to a major deviation from 48*180. Definitely not an increase of iops close to an order of magnitude.
Thanks for the suggestions.
Henry
Jakub Wartak wrote:
>
> We need more data, bascially OS and storage (FC, iSCSI?), versions
(OS, Data
> ONTAP), LVM configuration and NetApp model/configuration (sysconfig
command
> from SSH on NetApp too), I/O scheduler used if this is Linux, vmstat &
iostat
> outputs during the run, etc. NetApp has many optimization settings
(like
> no_update_atime on vols), FlexScale & PAM & PAM-II flash accelerators
(not to
> mention NVRAM write acceleration...), was deduplication turned on,
etc.
>
> There are some shortcut commands that you should launch on filer that
could
> solve the mystery, these are "lun show -o -i 1 /vol/path/to/lun" and
"stats
> show disk:*:disk_busy". Second is going to show you how much really
disks are
> utilised.
>
> BTW: LUN can be only be inside WAFL volume, volume can be only on
single
> aggregate. So my understanding is that LUN is in single aggregate that
has 3
> raid group sets each consisting of 16 disks, correct? As far as I
remember
> the calculation of IOPS in RAID-DP is not as simple as 48*180.
>
>
-- http://www.freelists.org/webpage/oracle-l -- http://www.freelists.org/webpage/oracle-lReceived on Tue May 04 2010 - 11:34:00 CDT