Re: [GSoC08]some detail plan of improving hash index

From: Josh Berkus <josh(at)agliodbs(dot)com>
To: pgsql-hackers(at)postgresql(dot)org
Cc: "Xiao Meng" <mx(dot)cogito(at)gmail(dot)com>, "Jonah H(dot) Harris" <jonah(dot)harris(at)gmail(dot)com>
Subject: Re: [GSoC08]some detail plan of improving hash index
Date: 2008-05-16 21:25:50
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-hackers


> Phase 1 seems extremely easy. I'm trying to do it first.
> Additionally, I need a benchmark to test the performance. It seems
> there's some tools list in
> . Any advice?

For a simple test, pgbench is actually going to be pretty good for hash
index since it's mostly primary key access. You also might want to write
your own unit tests using pgunittest, because you want to test the

bulk load, both COPY and INSERT
single-row updates, inserts and deletes
batch update by key
batch update by other index
batch delete by key
batch delete by other index
concurrent index updates (64 connections insert/deleting concurrently)

You can compare all of the above against b-tree and unindexed columns.

For a hard-core benchmark, I'd try EAStress (SpecJAppserver Lite)


Josh Berkus
PostgreSQL @ Sun
San Francisco

In response to


Browse pgsql-hackers by date

  From Date Subject
Next Message Merlin Moncure 2008-05-16 21:38:02 Re: libpq object hooks
Previous Message Tom Lane 2008-05-16 20:23:16 Re: libpq object hooks