Active - Active SRM Configuration

I am implementing SRM in a VDI environment and trying to configure it in an Active-Active Design to minimize on idle hardware, while maximizing the utilization of all available resources.  To accomplish the Active-Active Design, we are using shared storage with multiple arrays, and dedicating space for the images, while reserving space for replication, and vica-versa on the other array. My problems start in relation to the View environment and the connection brokers.
While manually provisioning and balancing machines across both sites won't be a huge chore, directing users to the sites with their Virtual desktop will be the challenge. Because we are attempting the Active-Active Design, I will have two View portals in the environment (Primary and Recovery) and I don't want to leave determining which site to go to up to the users. Ideally I would like to find a way to consolidate the View Portals into one URL using DNS, which can be done, BUT.. if the Primary site responds to the URL request and that User does not have a Virtual Machine on the Primary site, the log in will fail.. UNLESS, I can find a way to query both the Primary and Recovery site during the authentication process and have it then direct the user to the portal they have access to.  So the question becomes.. Does anyone have any insight, direction, Hints, or Tips on how this can be addressed? Is anyone out there doing anything similar ? Your time and help is appreciated.
Who is Participating?
za_mkhConnect With a Mentor Commented:
Not really knowing much about SRM, and your exact topology, can I offer the following advice in the hope that it helps?
Since the View desktops are going to be provisioned amongst multiple sites (SITEA and SITEB) at any given time, I assume that vCenter will only running at one site but managing machines across both sites. Since the View Manager Server connects to once vCenter instance, vCenter would know where desktop is running and automatically redirect the user to the correct desktop? I could be wrong here. But my understanding of this is the View Broker talks to vCenter to find ESX Host location of VM when a VMView session is initiated. Then a direct RDP/PCoIP session is established to the VM's IP address. In our environment, vCenter could move my virtual desktop to another ESX host but things continue to work while I am connected to it ... I'm not even aware that my desktop has moved from one part of our building to another. But of course, in our case, the VM's IP configuration remains intact since our VLANs run in both buildings.
DNS looks like the solution - with a static solution. In that you only have one record for your view server. So by default, the primary DNS record would be that of your primary site. In case of failure, you then update record so that it points to recovery site broker. Also if you use Microsoft DNS, the DNS server can be configured to return the closest IP address that matches the IP of the client making the DNS request. So this could help you too.
Another thing you can try, is to put a load balancer in front of your view connection manager servers. Maybe that could help.
MarcLamannaAuthor Commented:
Thanks for the comments, the solution is going to have to include a homegrown front-end that will act as a gateway for the View Brokers, and queries the VC Database for which SRM site they belong to. The DNS solution you mentioned will work, but the Homegrown Gateway is the only way to present an authentication window which queries the DB in the backend.
Thanks again for the help,
MarcLamannaAuthor Commented:
While the comments were helpful, I have come to the conclusion that the only solution will be a homeGrown Gateway type front-end.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.