From: "John Poltorak" Received: from mxout3.mailhop.org ([63.208.196.167] verified) by 2rosenthals.com (CommuniGate Pro SMTP 5.0.9) with ESMTP id 295190 for os2-wireless_users@2rosenthals.com; Thu, 17 Aug 2006 06:32:11 -0400 Received: from mxin1.mailhop.org ([63.208.196.175]) by mxout3.mailhop.org with esmtp (Exim 4.51) id 1GDfAC-0009mH-5o for os2-wireless_users@2rosenthals.com; Thu, 17 Aug 2006 06:31:48 -0400 Received: from 213-152-37-93.dsl.eclipse.net.uk ([213.152.37.93] helo=mail.warpix.org) by mxin1.mailhop.org with smtp (Exim 4.51) id 1GDfAB-00096g-Sf for os2-wireless_users@2rosenthals.com; Thu, 17 Aug 2006 06:31:48 -0400 Received: from tp600.warpix.org by mail.warpix.org (IBM OS/2 SENDMAIL VERSION 2.03/2.0) id LAA051.39; Thu, 17 Aug 2006 11:31:46 +0100 Received: by tp600.warpix.org (IBM OS/2 SENDMAIL VERSION 2.03/2.0) id LAA001.81; Thu, 17 Aug 2006 11:30:31 +0100 Date: Thu, 17 Aug 2006 11:30:30 +0100 To: "Rick R." Subject: Re: [OS2Wireless]Re: AFINETK crash - Dump Screen Message-ID: <20060817113030.B47@warpix.org> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.94.15i In-Reply-To: ; from Rick R. on Thu, Aug 17, 2006 at 03:18:28AM -0700 X-Mail-Handler: MailHop by DynDNS X-Spam-Score: -2.4 (--) On Thu, Aug 17, 2006 at 03:18:28AM -0700, Rick R. wrote: > Elementary that has NOTHING to do with the problem at hand. > And dear Mr. Watson, before ASSUMING that I didn't check this out over and over again you should know that assumption is the mother of all screw ups. > > I checked and tripple checked this over and over again. > Down to the time stamps of all driver files involved. > > Each and every driver is patched to the level in the newstack.zip file and it still doesn't do no good. > I even tried removing the "k" version and go with the SOCKETS * AFINET only. > Didn't do no good either - so don't assume, it does no good either. > > After eliminating all other possibillities the only conclusion left is that the OS/2 TCP/IP stack has an inherent flaw that makes it unstable at high network throughput. > That was never properly fixed by IBM and by now never will be. Just how much traffic are you talking about? And how do you measure it? I'm interested since I have no idea how much traffic my server gets and would like to know. -- John