I know whar ports are used with an ePO 4.6 infrastructure (KB66797). I can´t find how the full communication process works between the clients and the ePO server. ¿Do the clients connect with the ePO server or do they connect with the distributed repositories and those ones send all properties and info to the console? I would like to find that info so I know what ports should I open in the firewall or keep them closed.
As far as i know a client doesn't commuincate, in the way you described, with a distributed repository unless it 's a super agent.
The only comms made to a normal repo would based on it's type UNC/FTP/HTTPMessage was edited by: Tristan on 14/09/12 14:57:16 IST
thanks for quick reply. I forgot to mention that our distributed repositories will be superagents. I should have asked if the clientes communicate directly with the ePO server or with the superagents.
Directly with ePO: they will only contact the superagent machines if they are configured to use them as repositories (which is the most common use of SA machines.)
yes, the client agents will be configured to use the superagents as repositories where download the updates from. Does it mean so, that no communication will be performed within the client agent and the ePO server as they will connect to the superagents and those ones will send the properties to the server and viceversa?
Does it mean so, that no communication will be performed within the client agent and the ePO server as they will connect to the superagents and those ones will send the properties to the server and viceversa?
No, the opposite - the agents will talk directly to the ePO server for everything except software updates and deployments, which come from the superagents.
The agent get's it's software and updates from a repository server / SuperAgent, but for reporting and downloading policies the agent will connect directly to your ePO server (or Agent Handler).
Don't forget a SuperAgent is used to broadcast agent wake-up calls from the ePO server to the broadcast segment the SuperAgent is on.
So in essense there is an ePO -> SuperAgent -> Many Agents -> ePO conversation
True, but only if you use global updating (or send superagent wakeup calls.) You don't need to use these functions if you want the SA to function simply as a repository.