Re: Get explain output of postgresql in Tables - Mailing list pgsql-hackers

From Greg Sabino Mullane
Subject Re: Get explain output of postgresql in Tables
Date
Msg-id 4d2aae4b48cfe29ad3307f939f1ca9bf@biglumber.com
Whole thread Raw
In response to Re: Get explain output of postgresql in Tables  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Get explain output of postgresql in Tables  (Mischa Sandberg <mischas@ActiveState.com>)
List pgsql-hackers
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


> I wonder if it would help much just to change EXPLAIN to indent with
> something other than spaces?

I like that. Maybe even decrease the indenting a little more, and compress
some of the inner whitespace (such as the 2 spaces after the operator name)

One other thing I've done in the past that helps a lot is to simplify the text
by using "L" for loops, "W" for width, "C" for cost, and "R" for rows, and
even "AT" for "actual time".

This ends up saveing an enormous amount of horizontal screen space, and
is a really easy intuitive one-time learning curve.

Normal verbose way:
Sort  (cost=11383.82..11383.83 rows=1 width=38) (actual time=18942.712..18942.741 rows=9 loops=1)  Sort Key: count(*)
-> HashAggregate  (cost=11383.80..11383.81 rows=1 width=38) (actual time=18942.581..18942.612 rows=9 loops=1)        ->
Bitmap Heap Scan on turnstep_mail  (cost=134.73..11383.79 rows=1 width=38) (actual time=17085.967..18941.677 rows=193
loops=1)


Tom + Greg style:

Sort (C=11383.82..11383.83 R=1 W=38) (AT=18942.712..18942.741 R=9 L=1)
- -Sort Key: count(*)
- -->HashAggregate (C=11383.80..11383.81 R=1 W=38) (AT=18942.581..18942.612 R=9 L=1)
- ---->Bitmap Heap Scan on turnstep_mail (C=134.73..11383.79 R=1 W=38) (AT=17085.967..18941.677 R=193 L=1)

I use  capital letters as it makes it easier to read, especially for things like the common
single loop (L=1 vs. l=1)

- --
Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200604121213
http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8

-----BEGIN PGP SIGNATURE-----

iD8DBQFEPSkQvJuQZxSWSsgRAsc3AKDEWkJR6hHr2/Rgwgk49UNhGVtR6ACgo91Z
7Ck46wiCWoVvGW6V/AR7wAo=
=UKnc
-----END PGP SIGNATURE-----




pgsql-hackers by date:

Previous
From: Martijn van Oosterhout
Date:
Subject: Re: Practical impediment to supporting multiple SSL libraries
Next
From: "Merlin Moncure"
Date:
Subject: Re: plpgsql by default