Re: Patch: add timing of buffer I/O requests
Tom Lane <tgl@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Peter Geoghegan <peter@2ndquadrant.com>, Ants Aasma <ants@cybertec.at>, Greg Smith <greg@2ndquadrant.com>, pgsql-hackers@postgresql.org
Date: 2012-04-28T16:49:44Z
Lists: pgsql-hackers
Robert Haas <robertmhaas@gmail.com> writes: > ... You might want to revisit the issue of how the new > columns in pg_stat_statements are named, as well. I am not sure I'm > happy with that, but neither am I sure that I know what I'd like > better. It's not too clear that the timing is specifically for data > block reads and writes, for example. Well, the names "time_read" and "time_write" are certainly out of step with every other stats view in the system; everyplace else, such columns are named "something_time" (and even in this view itself the other timing column is "total_time", not "time_total"). So that's got to change. We could just reverse the word order to "read_time" and "write_time", or we could do something like "buf_read_time" or "data_read_time". IIUC block_read_time/block_write_time in the pg_stat_database view are database-wide totals for the same numbers, so perhaps the pg_stat_statements column names should be consistent with those. I am kinda wondering though why those columns spell out "block" where every single other column name in the stats views uses the abbreviation "blk". regards, tom lane