Forum Discussion

brent112_11716's avatar
brent112_11716
Icon for Nimbostratus rankNimbostratus
Mar 22, 2012

Exchange 2010 and BigIP v11.1??

Does anyone have a successful setup with BigIP V11.1 and Exchange 2010 (with multiple front end CAS servers)? We have had alot of issues since setting it up. The main issue is authentication prompts that Outlook users get, and when a server is rebooted that a users Outlook is pointed to, instead of failing over to another server nicely they get a nice authentication prompt from within Outlook. I have had a ticket opened with F5 since Mid December and while they have worked hard to help figure this out, the issue is still occurring. There is also an issue of sometimes Outlook clients getting a "server not available" when loading Outlook. The only solution is to reboot the computer and then everything works fine. If a computer is manually pointed (via host file) to an Exchange server directly there are 0 issues, only when they go through the F5 do we get these problems.

 

 

I was wondering if someone had a similar setup that is working without issue so I could compare some settings and such to see if I can get this working.

 

 

 

Thank's

 

Brent

 

  • mikeshimkus_111's avatar
    mikeshimkus_111
    Historic F5 Account
    The solution to this issue was to apply the same source IP persistence profile from the RPC virtual server to the virtual server receiving EWS traffic, and set that profile to match across virtual server.
  • Hi Mike,

     

    So you mean that all the Virtual Servers for Exchange 2010 services should use the same persistance profile... ???? which means if RPC VS will use the src_IP persistence then we need to change the persistence for the combined VS formed for HTTP based services as well, is it??????

     

     

    Regards,
  • mikeshimkus_111's avatar
    mikeshimkus_111
    Historic F5 Account
    That is incorrect. The combined virtual server assigns persistence within the iRule. In the instance where you had both a single virtual for HTTP services AND a virtual server for RPC clients, you would assign the source IP persistence profile from the RPC virtual as the fallback persistence profile on the HTTP virtual. Then, change the persistence method for EWS in the iRule to source IP. This will force EWS to persist according to the settings in the fallback profile, and send the EWS traffic from an RPC client to the same pool member as the RPC traffic.

     

     

    Clear as mud?