• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1780
  • Last Modified:

Connection to Microsoft Exchange has been restored. - SBS 2003 and Outlook (any version running on any OS)

For the last few months every single one of our Outlook clients drop and reconnect to the Exchange server whilst Outlook is running repeatedly, every minute.  

The only thing logged is client-side within event viewer:
Connection to Microsoft Exchange has been restored.
Connection to Microsoft Exchange has been lost. Outlook will restore the connection when possible.


The disconnects happen almost exactly every 60 seconds give or take a second or two.  The reconnect is virtually instantaneous.  This goes on forever but doesn't often start until 1 to 3 minutes have passed upon initially starting Outlook.

There is nothing logged on the Exchange server that I can find.

This happens whether Outlook is running in "cached mode" or not.  However when clients are not running in cached mode their experience is improved in the sense that messages seem to send quicker however the disconnects are still occurring.  

I can ping the Exchange server with 0% loss while the connection is dropped and reconnected.  

I can duplicate this at home on any of my desktops.  I can duplicate this in the office if I connect directly to our public WiFi (straight to the router and before the server and switch).  In other words, I can duplicate it anywhere on anything.

Here's what I've done:

* Replaced the switch
* Replaced the ethernet cables
* replaced the router
* replaced both NIC's
* everything in this article:
http://blogs.technet.com/b/sbs/archive/2007/04/24/common-networking-issues-after-applying-windows-server-2003-sp2-on-sbs.aspx
* Moved the Exchange transaction logs to their own hard drive

I'm at my wits end on this one and for some strange reason feel it's going to boil down to some obscure DNS problem that I can't figure out.  Please help and I'll name my next child after you!

Here's a screeny of a client's eventvwr
Event Viewer on Client
Here's a screeny of Outlook's connection status tool:
Outlook connection status
0
xfungalx
Asked:
xfungalx
  • 6
  • 4
1 Solution
 
xfungalxAuthor Commented:
I should add that this only happens for our "RPC over HTTP" clients.  In other words, you only experience this if you're not on the LAN proper.
0
 
TalkExchangeCommented:
I checked this
http://social.technet.microsoft.com/Forums/en-US/exchangesvrclients/thread/303efcdc-7720-46fa-a00c-0219e6c6725c/

Should be helpful, did you check for event logs for 9646 event ids.

0
 
xfungalxAuthor Commented:
I have seen that post and have run through its suggestions.  I have never seen that eventID logged while a remote Outlook user is connected and experiencing the problem.  I have also replaced the Broadcom NIC card that came with the server with two, new Intel boards with updated and current drivers.  

While I type this from home I am seeing the "Outlook has lost connection..." balloon pop up every minute, only to be met with "The connection to Microsoft Exchange has been restored" immediately afterwards.  I remote'd into the server just now just to check again and see no entries of 9649.

Maddening!

0
Free learning courses: Active Directory Deep Dive

Get a firm grasp on your IT environment when you learn Active Directory best practices with Veeam! Watch all, or choose any amount, of this three-part webinar series to improve your skills. From the basics to virtualization and backup, we got you covered.

 
TalkExchangeCommented:
how about starting outlook in safe mode. Does this too give you popups.
0
 
xfungalxAuthor Commented:
Good question, I just fired it up in safe mode and I'll check back on it in a little while and let you know.  
0
 
xfungalxAuthor Commented:
Sadly, it's the same result in safe mode.
0
 
TalkExchangeCommented:
Collect a netmon and check where the packet gets dropped.
0
 
TalkExchangeCommented:
Although this isnt related but you can check

http://blogs.technet.com/b/exchange/archive/2008/05/02/3405434.aspx

to check for few more things.

Do you have wan accelator for instance.
0
 
xfungalxAuthor Commented:
I found finally figured out the problem.  

Turns out it was the "Connection timeout" setting for the "Default Web Site" in IIS.  It was set to 30 seconds.  Ramping it up to 240 seconds has completely eliminated the issue.  120 is the default setting and I don't know why or how it got lowered to 30.  Presumably it was done to reduce the server load (RAM usage is up nearly 25% with the 240 second setting).  

Either way, I'm still doing fine as far as server resources go but this setting was most certainly the cause.  I can easily re-duplicate the issue by manipulating the timeout setting.  Not to mention it makes perfect sense in hindsight and explains why there were zero errors logged while this was going on.  As far as Exchange was concerned everything was "working as intended."

Hopefully someone can benefit from this if they too find themselves in this frustrating quandary.
0
 
xfungalxAuthor Commented:
Closing this issue and accepting my own solution as the fix.  

My thanks to everyone that posted, while none of the comments were the solution it did help me to eliminate possible causes and ultimately led me to a all things "timeout" related.
0

Featured Post

Prep for the ITIL® Foundation Certification Exam

December’s Course of the Month is now available! Enroll to learn ITIL® Foundation best practices for delivering IT services effectively and efficiently.

  • 6
  • 4
Tackle projects and never again get stuck behind a technical roadblock.
Join Now