Re: Reliability recommendations - Mailing list pgsql-performance

From Dan Gorman
Subject Re: Reliability recommendations
Date
Msg-id 125CA0A8-6142-4DF5-822E-CF0B0C46B4C1@hi5.com
Whole thread Raw
In response to Re: Reliability recommendations  (Mark Kirkwood <markir@paradise.net.nz>)
Responses Re: Reliability recommendations
Re: Reliability recommendations
List pgsql-performance
All,

Was that sequential reads? If so, yeah you'll get 110MB/s? How big
was the datafile size? 8MB? Yeah, you'll get 110MB/s. 2GB? No, they
can't sustain that. There are so many details missing from this test
that it's hard to have any context around it :)

I was getting about 40-50MB/s on a PV with 14 disks on a RAID10 in
real world usage. (random IO and fully saturating a Dell 1850 with 4
concurrent threads (to peg the cpu on selects) and raw data files)

Best Regards,
Dan Gorman

On Feb 24, 2006, at 4:29 PM, Mark Kirkwood wrote:

> Luke Lonergan wrote:
>
>> I'd be more shocked if this weren't also true of nearly all SCSI
>> HW RAID
>> adapters of this era.  If you had ordered an HP DL380 server you'd
>> get about
>> the same performance.
>> BTW - I don't think there's anything reasonable about 50-55 MB/s
>> from 6
>> disks, I'd put the minimum for this era machine at 5 x 30 = 150MB/s.
>
> He was quoting for 6 disk RAID 10 - I'm thinking 3 x 30MB/s = 90MB/
> s is probably more correct? Having aid that, your point is still
> completely correct - the performance @55MB/s is poor (e.g. my *ata*
> system with 2 disk RAID0 does reads @110MB/s).
>
> cheers
>
> Mark
>
> ---------------------------(end of
> broadcast)---------------------------
> TIP 4: Have you searched our list archives?
>
>               http://archives.postgresql.org


pgsql-performance by date:

Previous
From: Mark Kirkwood
Date:
Subject: Re: Reliability recommendations
Next
From: Mark Kirkwood
Date:
Subject: Re: Reliability recommendations