Re: PostgreSQL pre-fork speedup
James Robinson <jlrobins@socialserve.com>
From: James Robinson <jlrobins@socialserve.com>
To: sdv mailer <sdvmailer@yahoo.com>
Cc: PostgreSQL Development <pgsql-hackers@postgresql.org>, Tatsuo Ishii <t-ishii@sra.co.jp>
Date: 2004-05-06T17:07:42Z
Lists: pgsql-hackers
Quick overview of the code for differences in TCP-on-the-frontend code is a call to setsockopt(..., TCP_NODELAY, ...) if the connection to the frontend is a TCP socket. Could this be producing pseudo-fragmentation, resulting in over-the-top context switches? Looks like pool_process_query() does a lot of little itty bitty writes to the frontend filedescriptor. What do you get if you comment out that block in child.c, around line 372? Either a faster system or a non-working one? ---- James Robinson Socialserve.com