| From: | Jonathan Vanasco <postgres(at)2xlp(dot)com> |
|---|---|
| To: | "pgsql-general(at)postgresql(dot)org general" <pgsql-general(at)postgresql(dot)org> |
| Subject: | Benching Queries |
| Date: | 2014-10-02 20:13:16 |
| Message-ID: | 78B2522B-7B2C-4C92-ABB2-4DE079FFACCC@2xlp.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
Does anyone have a good solution for benching queries under various conditions, and collecting the EXPLAIN data ?
I looked at pgbench, but it doesn't seem to be what I want.
My situation is this-
- For a given query, there are 3-5 different ways that I can run it.
- Each form of the query has a completely different execution plan and query time, often using different indexes.
- The same query runs differently on first query, vs subsequent queries (when the indexes/tables are already in memory).
My goal is to find an overall balance of query time (cold-start vs in-memory) and indexes (number of).
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andrus | 2014-10-02 20:49:34 | Creating index on concatenated char columns fails is Postgres 9 (regression) |
| Previous Message | José Luis Rando Calvo | 2014-10-02 17:08:10 | Re: Postgres tcp_keepalive_xxxx parameters. |