X-Original-To: pgsql-hackers-win32-postgresql.org@localhost.postgresql.org Received: from localhost (av.hub.org [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 4A859D85D7; Thu, 8 Sep 2005 21:11:07 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 64292-09; Fri, 9 Sep 2005 00:11:00 +0000 (GMT) Received: from yass.opencloud.co.nz (yass.opencloud.co.nz [203.79.85.162]) by svr1.postgresql.org (Postfix) with ESMTP id 4CD70D72ED; Thu, 8 Sep 2005 21:11:01 -0300 (ADT) Received: from 203-79-104-60.cable.paradise.net.nz ([203.79.104.60] helo=[192.168.0.20]) by yass.opencloud.co.nz with asmtp (TLS-1.0:DHE_RSA_AES_128_CBC_SHA:16) (Exim 4.34) id 1EDWTv-000112-HH; Fri, 09 Sep 2005 12:11:03 +1200 Message-ID: <4320D317.10806@opencloud.com> Date: Fri, 09 Sep 2005 12:11:03 +1200 From: Oliver Jowett User-Agent: Mozilla Thunderbird 1.0.2 (X11/20050324) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Tom Lane Cc: Merlin Moncure , pgsql-patches@postgresql.org, pgsql-hackers-win32@postgresql.org Subject: Re: [PATCHES] Implement support for TCP_KEEPCNT, TCP_KEEPIDLE, TCP_KEEPINTVL References: <6EE64EF3AB31D5448D0007DD34EEB3417DD2C6@Herge.rcsinc.local> <4320C4BE.7010306@opencloud.com> <26590.1126221622@sss.pgh.pa.us> <4320CA2E.1000104@opencloud.com> <26710.1126222815@sss.pgh.pa.us> <4320CE22.4070408@opencloud.com> <26887.1126224143@sss.pgh.pa.us> In-Reply-To: <26887.1126224143@sss.pgh.pa.us> X-Enigmail-Version: 0.90.2.0 X-Enigmail-Supports: pgp-inline, pgp-mime Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.008 required=5 tests=[AWL=0.008] X-Spam-Level: X-Archive-Number: 200509/13 X-Sequence-Number: 3379 Tom Lane wrote: > I'm not sure if we can issue a notice that will be seen on the client > side at this point in the startup cycle. I seem to recall the protocol > document advising against sending NOTICEs during the authentication > cycle. As currently written the setsockopt() calls are done very early, well before even ProcessStartupPacket() has run -- so we can't really send anything at all to the client because we don't even know which protocol to use. Delaying the setsockopt() introduces the danger that you could lose the network at just the wrong time and have a backend sitting around for an extended period waiting on a startup packet (but not holding locks, admittedly). -O