locked
FQDN of the Web Component Server RRS feed

  • Question

  •  

    I am installing OCS 2007 in a lab environment.  The planning guide of the expanded deployment shows one load balancer in front of the Front End Servers and a second load balancer in front of the Web Component servers.  From this picture, I would expect that there would be one FQDN for the FE servers and a different FQDN for the Web Component Servers.  Is this correct?  Where/How does the FQDN of the web component servers get set?

     

    Thanks in advance,

    Mike

    Wednesday, August 27, 2008 9:09 PM

Answers

  • Hi,

    I think the problem is that during setup for the pool itself, you specified your pool name rather than the VIP for the web components server. There's no easy way to change this. My suggestion is to use wbemtest to fix it.

     

    https://blogs.pointbridge.com/Blogs/mcgillen_matt/Pages/Post.aspx?_ID=36

     

    This will give you the chance to reset the internal URLs for all those webcomponents services to be the internal FQDN of the VIP for your web components servers.

     

    Regards,

    Matt

     

     

    Tuesday, September 2, 2008 9:16 PM

All replies

  •  

    The setup program installs the files for the web compontents on your IIS servers and then you manually request certificates from IIS.  Setup also allows you to activate your web components server which adds it to the OCS global configuration for your forest so the FE servers know about the web component server or web farm.

    Look at section 3.5.2: http://technet.microsoft.com/en-us/library/bb663624.aspx

     

    Thursday, August 28, 2008 5:51 AM
  • I tried this but I am missing something (likely something obvious).  I install and activate the web components.  Then I go to validate the installation.  I get the following failure.

     

    URL: https://lcsp1.iad.nwa.com/GroupExpansion/Int/service.asmx
    Internal Error: ConnectFailure

     

    I know that the problem is that lcsp1.iad.nwa.com is pool name and it is pointed at the VIP for the load balancer on the FE server pool.  Should the FQDN of the pool be pointing at the vip of the FE and the vip of the Web Components?  The deployment guide has a diagram that I am trying to follow unfortunately, it does not include FQDN's for the load balancers.  I am modeling after "High Availability Enterprise Deployment Supporting IM and Conferencing for Internal Users only.

     

    Any help would be appreciated.

     

    Thursday, August 28, 2008 2:19 PM
  •  

    The FQDN pointing to the web components' NLB VIP should be different than that of the FE pool VIP.  The certs on the web servers should have thier own FQDN and then the VIP FQDN in thier SAN field.  The NLB should not be doing any SSL offloading.
    Thursday, August 28, 2008 4:41 PM
  • OK.  That's what I thought.  I have found the root of the problem. If I go to the Web Components Property page and check the Group Expansion tab, the Internal URL is set to the FQDN of the pool.  When should that have been set to the FQDN of the IIS VIP?  How do I change it now?

     

    I believe that the URL for the office address book and the web conferencing virtual server should also be updated.  Are they set in the same way?

     

    Thanks

    Thursday, August 28, 2008 6:34 PM
  • Hi,

    I think the problem is that during setup for the pool itself, you specified your pool name rather than the VIP for the web components server. There's no easy way to change this. My suggestion is to use wbemtest to fix it.

     

    https://blogs.pointbridge.com/Blogs/mcgillen_matt/Pages/Post.aspx?_ID=36

     

    This will give you the chance to reset the internal URLs for all those webcomponents services to be the internal FQDN of the VIP for your web components servers.

     

    Regards,

    Matt

     

     

    Tuesday, September 2, 2008 9:16 PM