DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma;
	h=mime-version:in-reply-to:references:date:message-id:subject:from:to
	:cc:content-type:content-transfer-encoding;
	b=CmvK24vUz5M3mhjqvezzuZUhNsvrkSLtANJYi5WQTQk3mh008ihveKiDH75kIk/C7t
	3I3UHYfaRmiTbIErBmGFhJez8mSk7rtIg/h+nNd9KClobU9zz7hG8zM1d9ABDKrnpE90
	v/gjfVJgkTSpwmwxz+WXGW+VwBw1OVniHonQU=
MIME-Version: 1.0
In-Reply-To: <4BD8DCEA.2020003@2ndquadrant.com>
References: <i2pf2fd819a1004281127sf91d988cua1d95d13d09dae09@mail.gmail.com>
	<4BD8DCEA.2020003@2ndquadrant.com>
Date: Wed, 28 Apr 2010 21:45:06 -0700
Message-ID: <u2zf2fd819a1004282145wc60c3b39o7f0679fec9f607f8@mail.gmail.com>
Subject: Re: more 10K disks or less 15K disks
From: Anj Adu <fotographs@gmail.com>
To: Greg Smith <greg@2ndquadrant.com>
Cc: pgsql-admin@postgresql.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Thanks Greg

We do not archive the WALs. We use application-level replication to
achieve redundancy. WAL archiving was difficult to support with the
earlier hardware we had ( 6x300G 10K disks Dell 2850) given the
volumes we were dealing with. The RAID card should be from the same
manufacturer (LSI in Dell's case).

On our current system (6 x 450G 15K disks), we are able to sustain 100
million records/day but probably not more.

On Wed, Apr 28, 2010 at 6:12 PM, Greg Smith <greg@2ndquadrant.com> wrote:
> Anj Adu wrote:
>>
>> I am faced with a =A0hardware choice for a postgres data warehouse
>> (extremely high volume inserts..over 200 million records a day)
>
> That's an average of 2314 per second, which certainly isn't easy to pull
> off. =A0You suggested you're already running this app. =A0Do you have any=
 idea
> how high the volume of data going to the WAL is relative to everything el=
se?
>
>> 12 x 600G disks (15K) =A0(the new Dell Poweredge C server)
>> or
>> 24 x 600G =A0(10K disks)
>>
>
> You can expect the 15K disks to be 30% (more sequential work) to 50%
> (random/commit work) faster than a similar 10K drive. =A0So from most
> perspectives, twice as many 10K drives should be considerably faster. =A0=
The
> main point of concern here is the commit rate, which you can't necessaril=
y
> improve just by throwing drives at it. =A0That's based on how fast the dr=
ives
> spin, so there's the potential to discover a 50% regression there compare=
d
> to the setup you have right now. =A0With the RAID card in there, it shoul=
d be
> fine, but it's something to be concerned about.
>
> Also, you didn't mentioned the RAID card for the new system, and whether =
it
> would be the same in both setups or not. =A0That can be as important as t=
he
> drives when you have larger arrays. =A0The LSI Megaraid card Dell is usin=
g for
> the Perc6i is quite fast, and you'll need to make sure you get something
> just as good for the new server.
>
> --
> Greg Smith =A02ndQuadrant US =A0Baltimore, MD
> PostgreSQL Training, Services and Support
> greg@2ndQuadrant.com =A0 www.2ndQuadrant.us
>
>