Re: smart or dumb partition?

From: Bob Gobeille <bob(dot)gobeille(at)hp(dot)com>
To: Greg Stark <gsstark(at)mit(dot)edu>
Cc: PgSQL-general <pgsql-general(at)postgresql(dot)org>
Subject: Re: smart or dumb partition?
Date: 2009-08-08 04:49:44
Message-ID: 6003BD76-E371-43AB-B37F-B32D0731A2F8@hp.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general


On Aug 7, 2009, at 5:44 PM, Greg Stark wrote:

> On Sat, Aug 8, 2009 at 12:27 AM, Bob Gobeille<bob(dot)gobeille(at)hp(dot)com>
> wrote:
>> I gather from rtfm that it is typical to set up partitions so that
>> the
>> "master" table has no records. But from my understanding of
>> partitions and
>> doing some tests, I don't see any reason that has to be. So I'm
>> wondering
>> if I'm missing some subtle (or not so subtle) point about partitions?
>
> It's purely a convenience issue. Any child partition can be removed
> later, The parent will be a pain if you ever want to remove it from
> the partitioning structure.

Cool. In my case this won't be an issue. My app loads in archives
(.tar, .iso, etc) and stores the contained file hierarchies in the db
(organized by the archive). So if an archive only has a few thousand
files, I'll just save those few thousand records in the master table.
But if the archive has millions of files (some distro iso's have ~7M
files), I want to store them in their own partition.

Many thanks for the confirmation!

Bob Gobeille
bobg(at)fossology(dot)org

In response to

Browse pgsql-general by date

  From Date Subject
Next Message David 2009-08-08 07:34:29 Arrays and LIKE
Previous Message Tom Lane 2009-08-08 00:35:57 Re: Make check fails on 8.3.7