I think I know the answer to this but I want to get some consensus on it.
When I click within the console and click “Force Refresh”, is the UDP packet to force a client to reply coming from the core server or is it being routed to the relay to be sent to the client?
It is routed through relay only.Until the relay is acting as mediator.
If the agent directly reporting to server than it is server.
Hope this helps !!
That’s what I figured. Just needed to see someone else say it.
One additional detail to consider here: the notification mechanism from Server -> Relay, and Relay -> Relay is via TCP (default port 52311). It is only UDP (default port 52311) from Relay -> Client. For reference, please see https://www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli%20Endpoint%20Manager/page/Network%20Traffic%20Guide (and if it is not terribly clear on this point, perhaps I can make some updates).
Not sure your use case but I had to do this for machines which showed any property in a state.
In addition to performing this action in the console from an on demand basis I also used this and scheduled it twice a day:
https://www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli%20Endpoint%20Manager/page/BES%20Client%20Refresher
Another option is to create a task to perform an action of “notify client ForceRefresh” and deploy it out so if the client is able to check it will perform forced refresh without relying on a UDP message.
As an aside (and if my memory serves me correctly) the client installed on the relay itself receives the UDP ping from itself. This means you don’t need UDP from the server to your relays at all – just TCP 52311.