Re: Hardware suggestions for maximum read performance

Note that with linux (and a few other OSes) you can use RAID-1E
http://en.wikipedia.org/wiki/Non-standard_RAID_levels#RAID_1E
with an odd number of drives.

On Fri, May 3, 2013 at 12:16 AM, Arjen van der Meijden
<acmmailing@tweakers.net> wrote:
> 3x200GB suggests you want to use RAID5?
>
> Perhaps you should just pick 2x200GB and set them to RAID1. With roughly
> 200GB of storage, that should still easily house your "potentially
> 10GB"-database with ample of room to allow the SSD's to balance the writes.
> But you save the investment and its probably a bit faster with writes
> (although your raid-card may reduce or remove the differences with your
> workload).
>
> You can then either keep the money or invest in faster cpu's. With few
> concurrent connections the E5-2643 (also a quad core, but with 3.3GHz cores
> rather than 2.4GHz) may be interesting.
> Its obviously a bit of speculation to see whether that would help, but it
> should speed up sorts and other in-memory/cpu-operations (even if you're not
> - and never will be - cpu-bound right now).
>
> Best regards,
>
> Arjen
>
>
> On 3-5-2013 1:11 Mike McCann wrote:
>>
>> Hello,
>>
>> We are in the fortunate situation of having more money than time to help
>> solve our PostgreSQL 9.1 performance problem.
>>
>> Our server hosts databases that are about 1 GB in size with the largest
>> tables having order 10 million 20-byte indexed records. The data are
>> loaded once and then read from a web app and other client programs.
>>   Some of the queries execute ORDER BY on the results. There are
>> typically less than a dozen read-only concurrent connections to any one
>> database.
>>
>> SELECTs for data are taking 10s of seconds. We'd like to reduce this to
>> web app acceptable response times (less than 1 second). If this is
>> successful then the size of the database will grow by a factor of ten -
>> we will still want sub-second response times.  We are in the process of
>> going through the excellent suggestions in the "PostgreSQL 9.0 High
>> Performance" book to identify the bottleneck (we have reasonable
>> suspicions that we are I/O bound), but would also like to place an order
>> soon for the dedicated server which will host the production databases.
>> Here are the specs of a server that we are considering with a budget of
>> $13k US:
>>
>>     HP ProLiant DL360p Gen 8
>>     Dual Intel Xeon 2.4GHz 4-core E5-2609 CPUs
>>     64GB RAM
>>     2x146GB 15K SAS hard drives
>>     3x200GB SATA SLC SSDs
>>     + the usual accessories (optical drive, rail kit, dual power supplies)
>>
>> Opinions?
>>
>> Thanks in advance for any suggestions you have.
>>
>> -Mike
>>
>> --
>> Mike McCann
>> Software Engineer
>> Monterey Bay Aquarium Research Institute
>> 7700 Sandholdt Road
>> Moss Landing, CA 95039-9644
>> Voice: 831.775.1769  Fax: 831.775.1736 http://www.mbari.org
>>
>
>
> --
> Sent via pgsql-performance mailing list (pgsql-performance@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-performance



--
To understand recursion, one must first understand recursion.