Jim Lawrence
accessd at shaw.ca
Sun Mar 18 11:21:44 CDT 2012
Hi John: I just dug this out of the link you sent. There are three posts being combined but I think thye make it clear the solution they found: " The problem seems to be that the one or both servers are auto negotiating proper GB speed. The way large data is handled between the two platforms may have also changed somewhat and cards drivers are not setting proper spped. Focus on the network cards in each server. Go to the Network card properties of the card and select the configure button. Select the Advanced TAB and scroll through the options and look for LINK SPEED and DUPLEX, select Auto-negotiate 1000BPS. There may also be a choice auto-negotiate without the 1000BPS - this won't or may not work and that is the problem. Once we make this change we see the server transfer speed run properly. This fixed it for me but you said you already tried that. Try going into the network adapter driver properties and DISABLE "TCP Connection Offload (IPv4)". This is generally listed on the advanced tab of the network adapter. I have a fix that worked which worked perfectly. I have an HP ML 350 G6 server that was newly installed today and experience the same problem. I noticed that the HP software was not installed as I would have expected on older servers/oses. I downloaded the latest PSP from the HP drivers and downloads, installed it as a proliant server should and voila - the NIC and network worked exactly as it should. Before: Win 2003 SBS server to Win 2008 Enterprise w SP2 32bit -- was about 2 minutes for a 22MB folder After: Now it is literally 1 second !!! No tweaking needed - just installed and configured the HP Proliant software to SPEC !!" How this solves your problem. Jim -----Original Message----- From: accessd-bounces at databaseadvisors.com [mailto:accessd-bounces at databaseadvisors.com] On Behalf Of jwcolby Sent: Sunday, March 18, 2012 5:22 AM To: Access Developers discussion and problem solving Subject: Re: [AccessD] OT: Network speeds Yep. There is a switch at either end (behind the servers) and a switch in the middle. But is the a switch the issue at all? I am reading a ton of "Windows 2008 R2 to anything transfer speed sucks" stuff out on Google. For example: http://social.technet.microsoft.com/Forums/en-US/windowsserver2008r2general/ thread/2e14aa58-cb9b-462c-8a9e-7b10be2de3cb/ The servers at both end are Windows 2008 R2. It looks like my work is cut out for me. John W. Colby Colby Consulting Reality is what refuses to go away when you do not believe in it On 3/17/2012 11:40 PM, Stuart McLachlan wrote: > There are all sorts of possibilities there. What sort of rates do you get: > > 1. Through Switch 1 alone between SQL Server (computer) Win 2008 x64 and Unraid file > server (computer) > > 2. Through Switches 1 and 2 between SQL Server (computer) Win 2008 x64 and Living room > public PC (computer) > > 3. Through Switches 1 and 2 between Unraid file server (computer) and Living room public > PC (computer) > > 4. Through Switches 2 and 3 between Living room public PC (computer) and VM Server (computer) Win 2008 X46 > > 5. 4. Through Switches 2 and 3 between Living room public PC (computer) and Dev > workstation laptop (computer) > > 6. Through Switches 1,2 and 3 between Unraid file server (computer) and VM Server > (computer) Win 2008 X46 > > 7. Through Switches 1,2 and 3 between Unraid file server (computer) and Dev workstation > laptop (computer) > > 7. Through Switches 1,2 and 3 between SQL Server (computer) Win 2008 x64 and Dev > workstation laptop (computer) > > (I think that covers all the combinations!) > > An analysis of those results should tell you where the bottleneck is. > -- AccessD mailing list AccessD at databaseadvisors.com http://databaseadvisors.com/mailman/listinfo/accessd Website: http://www.databaseadvisors.com