| From: | Marten Feldtmann <marten(at)feki(dot)toppoint(dot)de> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | Mitch Vincent <mitch(at)huntsvilleal(dot)com>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Why Not MySQL? |
| Date: | 2000-05-03 17:57:30 |
| Message-ID: | 200005031757.TAA03161@feki.toppoint.de |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> "Mitch Vincent" <mitch(at)huntsvilleal(dot)com> writes:
>
> Well, there's still a cost to having a lot of seldom-used indexes,
> because the planner has to sit there and consider whether to use each
> one for each query. So I'd still recommend looking at your mix of
> queries and only creating indexes that match reasonably commonly-used
> WHERE clauses.
>
When doing insert/updates on larger tables (>500.000 entries) these
indexes are also time consuming !
For our vertical attribute object storage systems we noticed, that
the time for insert/updates are the critical part - they behave very
linear in our test suite and they seem to be the limiting factor in
our system.
Marten
| From | Date | Subject | |
|---|---|---|---|
| Next Message | The Hermit Hacker | 2000-05-03 18:03:27 | Re: 7.0RC2 compile error ! |
| Previous Message | Tom Lane | 2000-05-03 17:53:37 | Re: Why Not MySQL? |