Endpoint Protection

Hello,

It is possible for a Group Update Provider to fail over to a secondary server if it's primary management server is down?

Consider this:

1. GUP is getting definition updates from Server A
2. Server A goes down.
3. GUP unable to connect to server A
4. GUP connects to Server B (second server on the list)
5. GUP receives definition updates from Server B
6. Server A comes back
7. GUP starts updating from server A again.

Is this possible, and if so, has anyone made it work succesfully?

You need to configured the fail over for that.

Failover will cause a client to connect to another server on its management server list if the one it is currently connected to becomes unavailable.

In PUSH mode however, the client will not check periodically to see if the original manager is back up and switch to it. It continues to use the connection to the current SEPM until that connection is interrupted or the client reboots.

In PULL mode the client uses the priority listed in the Management Server List each time it checks into the manager because it does not maintain that communication as it would in PUSH mode.

But the failover will happen for all the clients. Its wont be the case that only the  GUP will failover to the new server.

Cheers,
Aniket

If the clients fail over as well, will GUP continue providing them with updates from the new server? I will use PULL technology.

As long as the LiveUpdate Settings policy tells them to, no matter which SEPM they connect to. 

The failover server will have the information about the groups and the policies on the first server using replication.When you apply a MSL to a group, the IP addresses specified in the MSL are present in the sylink.xml that the clients and the GUP has.

So, when the GUP will failover to the new server, it will have the IPs of both the servers in its sylink.xml.

The GUP should pull the definitions from the new server.

Cheers,
Aniket

If you configure a failover you must need a MS SQL database for better performance, you need to configure a replication partner and make a management list to have a failover. 

Thank you, my questions are answered.