| From: | Grzegorz Jaśkiewicz <gryzman(at)gmail(dot)com> |
|---|---|
| To: | Richard Broersma <richard(dot)broersma(at)gmail(dot)com> |
| Cc: | Merlin Moncure <mmoncure(at)gmail(dot)com>, John R Pierce <pierce(at)hogranch(dot)com>, PGSQL Mailing List <pgsql-general(at)postgresql(dot)org> |
| Subject: | Re: Avoiding surrogate keys |
| Date: | 2010-05-04 14:44:45 |
| Message-ID: | y2m2f4958ff1005040744o4a9b0deu58c0e23380a467fe@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
the rule of thumb for me is:
- if you have more than one column as PK - and are variable length,
or more than 2 columns, fixed length, no bigger than 8 bytes - go for
surrogate - always.
- if PK is variable length, on average longer than 8 bytes, or can
change - go surrogate.
- Otherwise leave it as it is.
At the end of a day, joining tables, using that have more than 1
variable length key, kills performance. In essence it makes index
search a PITA, and should be avoided.
This should always be however decision based on query performance,
usage, and data types. Never a textbook A or B type of thing.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andy Colson | 2010-05-04 14:50:28 | Re: pg9 beta1, make check fails |
| Previous Message | Tom Lane | 2010-05-04 14:37:43 | Re: pg9 beta1, make check fails |