From: | Douglas Trainor <trainor(at)uic(dot)edu> |
---|---|
To: | Tony Reina <reina_ga(at)hotmail(dot)com> |
Cc: | pgsql-admin(at)postgresql(dot)org |
Subject: | Re: Database storage bloat |
Date: | 2004-04-08 11:41:37 |
Message-ID: | 40753A71.3060104@uic.edu |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Saying "we've set field sizes to their theoretical skinniness" makes me
think that
you may have the wrong data types. For example, you may have used CHAR
and not VARCHAR.
douglas
Tony Reina wrote:
>I'm developing a database for scientific recordings. These recordings
>are traditionally saved as binary flat files for simplicity and
>compact storage. Although I think ultimately having a database is
>better than 1,000s of flat files in terms of data access, I've found
>that the database (or at least my design) is pretty wasteful on
>storage space compared with the binary flat files.
>
>In particular, I tried importing all of the data from a binary flat
>file that is 1.35 MB into a PostgreSQL database (a very small test
>file; average production file is probably more like 100 MB). The
>database directory ballooned from 4.1 MB to 92 MB (a bloat of 65X the
>original storage of the binary flat file).
>
>Now I know that table design and normalizing is important. As far as
>my partner and I can tell, we've made good use of normalizing (no
>redundancy), we've set field sizes to their theoretical skinniness,
>[...]
>
>
From | Date | Subject | |
---|---|---|---|
Next Message | Terry Hampton | 2004-04-08 12:36:37 | Location of a new column |
Previous Message | Gaetano Mendola | 2004-04-08 11:34:22 | CONTEXT: in log file |