| From: | Richard Huxton <dev(at)archonet(dot)com> |
|---|---|
| To: | Alban Hertroys <alban(at)magproductions(dot)nl> |
| Cc: | Sim Zacks <sim(at)compulab(dot)co(dot)il>, pgsql-general(at)postgresql(dot)org |
| Subject: | Re: understanding explain data |
| Date: | 2006-05-10 20:00:02 |
| Message-ID: | 44624642.1060104@archonet.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
Alban Hertroys wrote:
> Sim Zacks wrote:
>> Something such as: with this explain data, adding an index on table
>> tbl column A would drastically improve the efficiency. Or at least an
>> application that would say, the least efficient part of your query is
>> on this part of the code so that you could more easily figure out what
>> to do about it.
>
> The latter part is the most useful IMO, optimizing usually needs a
> (human) brain to put things into the right perspective. Adding an index
> can speed up your queries only so much, a more optimal data presentation
> (like moving calculations to insert/update instead of select) can do a
> lot more sometimes.
>
> It looks like something like that shouldn't be too hard to write...
> Maybe it even does exist already. Personally I'd prefer a command line
> tool ;)
> It would help if you can pipe the output of explain analyze to an
> external tool from within psql.
I've got a short perl script that I throw explain output into. It's not
brilliant - can give false positives, but it will usually give you two
or three lines to look at.
For those that are interested, I've attached it.
--
Richard Huxton
Archonet Ltd
| Attachment | Content-Type | Size |
|---|---|---|
| parse_explain.pl | application/x-perl | 554 bytes |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bob Pawley | 2006-05-10 22:04:23 | Re: Interface |
| Previous Message | Frank L. Parks | 2006-05-10 19:30:27 | Re: Interface |