none
UAG Expansion and Resiliency RRS feed

  • Question

  • UAG Expansion and Resiliency

    We are expanding our UAG environment from 3 standalone appliances in 3 global location to 8 appliances in 4 global locations. We will be enabling the array features of the paired appliances and implementing Global Dispatch across all 4 locations. In three of the locations we will be simply adding hardware and then creating the array using the existing configurations. In the 4ht the location we will be building an new implementation. Each of the existing UAG has between 10 and 15 portals, some with the client and some without, some with AD Authentication and some without.

    Does anyone have “Tips/Tricks/Gotchyas!” that I should be watching out for? Has anyone had any experience with Global Dispatch and if so, what is your experience with this? I have been using this product since Whale released 2.X and everything was manual, so I have a fairly healthy grasp on the product, what it can, and most importantly, can’t do.

    Any advice or comments are welcome.

    Monday, October 31, 2011 5:09 PM

All replies

  • I didn't have any Global Dispatch knowledge to share so I didn't repond, but it would be interesting to hear how you got on?


    Jason Jones | Forefront MVP | Silversands Ltd | My Blogs: http://blog.msedge.org.uk and http://blog.msfirewall.org.uk
    Friday, November 18, 2011 9:48 AM
    Moderator
  • Moving right along, we have expanded from 3 to 8 appliances, updated all the appliances to UAG SP1UP1RU1 and TMGSP2RU1.
    We have joined the domains, created the arrays and are now beginning to install
    and configure the Global Dispatch software to bolster our soft arrays. <o:p></o:p>

    We did test the hard array failover and have found it to be quite effective in the event of
    a hardware failure, but somewhat lacking if there is a soft issue. So, between
    NLB and Global Dispatch, we feel both scenarios should be adequately covered.<o:p></o:p>

    More to come...


    Monday, February 13, 2012 3:33 PM
  • Cool...good stuff

    Jason Jones | Forefront MVP | Silversands Ltd | My Blogs: http://blog.msedge.org.uk and http://blog.msfirewall.org.uk

    Tuesday, February 14, 2012 12:45 AM
    Moderator
  • ... and here we are at the close of the project.  This has been an excersise for sure, but one that has resulted in a more efficent and streamlined system.  We now run 8 appliances, each with about 15 portal, hosting almost 250 applications to almost 10k unique users around the world.  We use NLB for our array redundancy and Resonates Global Dispatch for our URL resiliency.  While we have provided a much more robust and resilient system to our clients, the end directive, we have created a monitoring monster from which there is no return.

    "Towards thee I roll, thou all-destroying but unconquering whale; to the last I grapple with thee; from hell’s heart I stab at thee; for hate’s sake I spit my last breath at thee. "  - Moby Dick, Herman Melville

    Tuesday, June 12, 2012 3:32 PM
  • Wow, impressive setup - what problems did you experience?

    P.S. I like the Whale reference, very clever :)


    Jason Jones | Forefront MVP | Silversands Ltd | My Blogs: http://blog.msedge.org.uk and http://blog.msfirewall.org.uk


    Tuesday, June 12, 2012 8:27 PM
    Moderator
  • So, we have dropped Global dispatch as our soft resiliency solution and are now evaluating TZO.  GD could not provide what we needed so we have had to look elsewhere.  

    As for issues, the migrations all went very smoothly, with the exception of the soft resiliency.  We are starting to see some fractures in one of our arrays and am not sure if its, UAG, TMG or the OS.  The only issue we have right now is service contention issues with UAG and IIS where W3WP and SessionMgrCom contend for resources and wind up locked.  MS says solution is in SP2 and private fix, but we can't push SP2 without a 60day lead time.  Also, installing SP2 on our preprod servers have shown issues in the installation and configuration migrations.(where neither work).

    Friday, August 24, 2012 2:20 PM