Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have several clients that will only connect to Exchange every other time they open outlook. It will wait around for the timeout period and then say "Failed to connect to the Exchange Server" (or something like that) and you have the option of Retry, Abort, and Cancel. If you retry it will work at least one of the next two retries. If you cancel and open outlook again it will connect as normal. Exchange is version 2003, clients are either 2003 or 2007, both have the same problem on affected PC's.

I updated the Broadcom driver on the Exchange box to the latest version, and have tried changing the timeout in outlook, neither worked. All other network connections are functioning normally on the affected machines, including always-on DB connections and a live terminal emulator connected to an AS/400.

Ideas? Fixes? Thanks.

edit: Tried rebuilding the outlook profile on one machine already.

edit2: More investigation shows that it is only certain users experiencing the problem, and those users only experience the problem when they are on a different subnet than the exchange server. Is it more likely an Exchange/AD authentication issue, or an issue with the routers for the separate subnets? Or both?

share|improve this question
Any errors in the Event Viewer on the server or clients? – Kyle Brandt Aug 17 '09 at 19:30
nothing relevant – therulebookman Aug 17 '09 at 19:49
I might also run a constant ping from the boxes, rule out network flakiness... – Kyle Brandt Aug 17 '09 at 19:51
not the problem...tried constant ping with large MTU size and got no drops. – therulebookman Aug 17 '09 at 20:09
Are you using RPC over HTTP? – DanBig Sep 3 '09 at 15:20

We very recently had this exact same problem, and the culprit was an overzealous PIX firewall and proxyarp. According to Cisco, this is a known issue. Check the client's arp entries for the Exchange server, and you will probably see the incorrect mac address in some cases.

We fixed this by issuing "sysopt noproxyarp inside" on our PIX. Obviously, make sure you understand the implications of disabling proxyarp, first.

share|improve this answer

First check to make sure the clients haven't exceeded their MAPI connection limit. Just like Kyle says the first thing to do is check for errors in event viewer.

If nothing shows up there, you can always try to deleting the .srs file, or if needed rebuild the outlook profile.

share|improve this answer
deleting the .srs didn't fix it either. – therulebookman Aug 18 '09 at 14:31
Are these clients on Laptops. – geeklin Aug 18 '09 at 19:12
no, not laptops – therulebookman Aug 19 '09 at 20:11

If users on the same LAN segment as the Exchange server don't experience the issue but users on a different LAN segment do experience the issue, then it's clear to me that the problem is related to the network. Check the switch in the affected LAN for congestion, errors, drops, etc. and check the routers that connect the LAN segments for congestion, errors, drops, etc.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.