View Connection Broker redundancy questions

Our environment currently is at about 100 users and quickly expanding.

My question has to do with the connection brokers.

We have two internal connection brokers. One is the primary that all our "internal" zero-clients are pointed to.

The second connection broker is a replica but it is paired with a "security" server in our DMZ to accomodate remote users.

So--we configured them so that internal clients are all pointed to the "primary" connection server "without" any SSL encryption aka "direct connect".

Remote user's coming in through the security server are routed through the second connection broker with SSL encryption enabled.

This setup was what seemed to be best practices according to VMware.

My question is if i want redundancy on our primary connection server do i just need to add another "replica" server?

If all our zero-clients are pointed to the IP of the primary connection server--will they automatically switch over to the replica connection broker if the primary fails?  Or is this a manual process.

Just wanting some input as I'm trying to see what everyone's doing as far as redundancy in the View environment.

Also any thoughts or recommendations as far as DR in a View environment would be appreciated.  WE have a DR sight but are still developing how we would "fail-over" our View environment as there are a lot of pieces to the puzzle.  I know a lot of it would be a manual process--but any input or suggestions would be appreciated.

Oh--and we are on View 4.6 currently.  We have 6 hosts running ESXi 4.1 U1 and a Dell Equallogic PS6500X SAN.


0 Kudos
2 Replies
VMware Employee
VMware Employee

>> My question is if i want redundancy on our primary connection server do i just need to add another "replica" server?

Yes, add one more replica server to the system. For redundancy, Configure Microsoft network load balencer and configure Hostname and IPddress for the Cluster Interface. Then login view administrator page  and configure Configure external URL for both the connection server replacing the hostname with cluster interface hostname. This FQDN should be resolvable from internal network directly. Client can connect to this external URL and will be Load balanced/or fail overed. If one connection fails, exisiting connection can re establish the connection using the same authentication.

0 Kudos


Check the following thread

Hope this information helps you.

0 Kudos