Re: Hardware/OS recommendations for large databases

From: Ron <rjpeace(at)earthlink(dot)net>
To: Alex Turner <armtuk(at)gmail(dot)com>,pgsql-performance(at)postgresql(dot)org
Subject: Re: Hardware/OS recommendations for large databases
Date: 2005-11-16 21:21:25
Message-ID: 6.2.5.6.0.20051116161052.03d54a70@earthlink.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Amendment: there are graphs where the 1GB Areca 1160's do not do as
well. Given that they are mySQL specific and that similar usage
scenarios not involving mySQL (as well as most of the usage scenarios
involving mySQL; as I said these did not follow the pattern of the
rest of the benchmarks) show the usual pattern of the 1GB 1160's in
1st place or tied for 1st place, it seems reasonable that mySQL has
something to due with the aberrant results in those 2 (IIRC) cases.

Ron

At 03:57 PM 11/16/2005, Ron wrote:
>You _ARE_ kidding right? In what hallucination?
>
>The performance numbers for the 1GB cache version of the Areca 1160
>are the _grey_ line in the figures, and were added after the
>original article was published:
>
>"Note: Since the original Dutch article was published in late
>January, we have finished tests of the 16-port Areca ARC-1160 using
>128MB, 512MB and 1GB cache configurations and RAID 5 arrays of up to
>12 drives. The ARC-1160 was using the latest 1.35 beta firmware. The
>performance graphs have been updated to include the ARC-1160
>results. Discussions of the results have not been updated, however. "
>
>With 1GB of cache, the 1160's beat everything else in almost all of
>the tests they participated in. For the few where they do not win
>hands down, the Escalade's (very occasionally) essentially tie.
>
>These are very easy to read full color graphs where higher is better
>and the grey line representing the 1GB 1160's is almost always
>higher on the graph than anything else. Granted the Escalades seem
>to give them the overall best run for their money, but they still
>are clearly second best when looking at all the graphs and the CPU
>utilization numbers in aggregate.
>
>Ron
>
>
>
>At 12:08 PM 11/16/2005, Alex Turner wrote:
>>Yes - that very benchmark shows that for a MySQL Datadrive in RAID 10,
>>the 3ware controllers beat the Areca card.
>>
>>Alex.
>>
>>On 11/16/05, Ron <rjpeace(at)earthlink(dot)net> wrote:
>> > Got some hard numbers to back your statement up? IME, the Areca
>> > 1160's with >= 1GB of cache beat any other commodity RAID
>> > controller. This seems to be in agreement with at least one
>> > independent testing source:
>> >
>> > http://print.tweakers.net/?reviews/557
>> >
>> > RAID HW from Xyratex, Engino, or Dot Hill will _destroy_ any
>> > commodity HW solution, but their price point is considerably higher.
>> >
>> > ...on another note, I completely agree with the poster who says we
>> > need more cache on RAID controllers. We should all be beating on the
>> > RAID HW manufacturers to use standard DIMMs for their caches and to
>> > provide 2 standard DIMM slots in their full height cards (allowing
>> > for up to 8GB of cache using 2 4GB DIMMs as of this writing).
>> >
>> > It should also be noted that 64 drive chassis' are going to become
>> > possible once 2.5" 10Krpm SATA II and FC HDs become the standard next
>> > year (48's are the TOTL now). We need controller technology to keep up.
>> >
>> > Ron
>> >
>> > At 12:16 AM 11/16/2005, Alex Turner wrote:
>> > >Not at random access in RAID 10 they aren't, and anyone with their
>> > >head screwed on right is using RAID 10. The 9500S will still beat the
>> > >Areca cards at RAID 10 database access patern.
>> > >
>> > >Alex.
>> > >
>> > >On 11/15/05, Dave Cramer <pg(at)fastcrypt(dot)com> wrote:
>> > > > Luke,
>> > > >
>> > > > Have you tried the areca cards, they are slightly faster yet.
>> > > >
>> > > > Dave
>> > > >
>> > > > On 15-Nov-05, at 7:09 AM, Luke Lonergan wrote:
>> > > >
>> > > >
>> > > >
>> > > >
>> > > >
>> > > > I agree - you can get a very good one from www.acmemicro.com or
>> > > >
>> > > > www.rackable.com with 8x 400GB SATA disks and the new 3Ware
>> 9550SX SATA
>> > > >
>> > > > RAID controller for about $6K with two Opteron 272 CPUs and 8GB of RAM
>> > > >
>> > > > on a Tyan 2882 motherboard. We get about 400MB/s sustained disk read
>> > > >
>> > > > performance on these (with tuning) on Linux using the xfs filesystem,
>> > > >
>> > > > which is one of the most critical factors for large databases.
>> > > >
>> > > >
>> > > >
>> > > >
>> > > > Note that you want to have your DBMS use all of the CPU and
>> disk channel
>> > > >
>> > > > bandwidth you have on each query, which takes a parallel database like
>> > > >
>> > > > Bizgres MPP to achieve.
>> > > >
>> > > >
>> > > >
>> > > >
>> > > > Regards,
>> > > >
>> > >
>> > >---------------------------(end of broadcast)---------------------------
>> > >TIP 2: Don't 'kill -9' the postmaster
>> >
>> >
>> >
>> >
>
>
>

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Claus Guttesen 2005-11-16 21:50:19 Re: Hardware/OS recommendations for large databases (5TB)
Previous Message Ron Mayer 2005-11-16 21:08:58 Re: Hardware/OS recommendations for large databases (