| From: | John Naylor <john(dot)naylor(at)enterprisedb(dot)com> |
|---|---|
| To: | Heikki Linnakangas <hlinnaka(at)iki(dot)fi> |
| Cc: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Force lookahead in COPY FROM parsing |
| Date: | 2021-03-18 15:09:34 |
| Message-ID: | CAFBsxsFQwDMmCcLCWuVPWX=SQvu0dtr6urPDV45sUUNVOC=MtA@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
The cfbot thinks this patch no longer applies, but it works for me, so
still set to RFC. It looks like it's because the thread to remove the v2
FE/BE protocol was still attached to the commitfest entry. I've deleted
that, so let's see if that helps.
To answer the side question of whether it makes any difference in
performance, I used the blackhole AM [1] to isolate the copy code path as
much as possible. Forcing lookahead seems to not make a noticeable
difference (min of 5 runs):
master:
306ms
force lookahead:
304ms
[1] https://github.com/michaelpq/pg_plugins/tree/master/blackhole_am
--
John Naylor
EDB: http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | David Steele | 2021-03-18 15:12:03 | Re: Implement <null treatment> for window functions |
| Previous Message | David Steele | 2021-03-18 15:04:19 | Re: Reduce the time required for a database recovery from archive. |