Re: Maximum Possible Insert Performance? - Mailing list pgsql-performance

From Josh Berkus
Subject Re: Maximum Possible Insert Performance?
Date
Msg-id 200311241004.37969.josh@agliodbs.com
Whole thread Raw
In response to Re: Maximum Possible Insert Performance?  (William Yu <wyu@talisys.com>)
Responses Re: Maximum Possible Insert Performance?  (William Yu <wyu@talisys.com>)
List pgsql-performance
William,

> The SanDisks do seem a bit pokey at 16MBps. On the otherhand, you could
> get 4 of these suckers, put them in a mega-RAID-0 stripe for 64MBps. You
> shouldn't need to do mirroring with a solid state drive.

I wouldn't count on RAID0 improving the speed of SANDisk's much.  How are you
connecting to them?  USB?   USB doesn't support fast parallel data access.

Now, if it turns out that 256MB ramdisks are less than 1/5 the cost of 1GB
ramdisks, then that's worth considering.

You're right, though, mirroring a solid state drive is pretty pointless; if
power fails, both mirrors are dead.

As I said before, though, we're all very interested in this test.  Using a
ramdisk for WAL has been discussed on this list numerous times but not
attempted by anyone who published their results.

All that aside, though, I think you should also experiment with the Background
Writer patch recently discussed on Hackers, as it may give you a performance
boost as well.
--
Josh Berkus
Aglio Database Solutions
San Francisco

pgsql-performance by date:

Previous
From:
Date:
Subject: Re: VACUUM problems with 7.4
Next
From: William Yu
Date:
Subject: Re: Maximum Possible Insert Performance?