When you send a wakeup call to these machines, you'll see that there is an option to use either "last connected agent handler" or "all agent handlers". If you leave it on "last connected", as long as the machine has not moved and the agent handler can see it - i.e it's not behind a NAT device, for example - then the wakeup call should succeed.
But isn´t that only an option to which AH the client should report back ?
Afaik the Wakeup-Call itself (same as agent push install) is done through the ePO Master.Nachricht geändert durch metalhead on 19.05.11 13:30:14 MESZ
No, both wakeup calls and agent installs can be carried out by agent handlers If you look at the configuration page for a send agent install, at the bottom you'll see that you can choose which handler to use to send the install - the wakeup call works in a similar way.
I would be very pleased if it works like that
I always interpreted "connect using" as the way the client will initially talk back to the server/ah, but not for deployment.
Re-reading the AH Whitepaper also shows your description:
"Agent handlers check the work queue frequently (approximately every ten seconds)
and perform the requested action. Typical actions include agent wake-ups, deployments, and
data channel messages."
Thanks for this one
I have short question: If the agent is outside of DMZ( mean is at the internet client is more than 1000) Which option do you recommend to use?! and what is best to do agent wakeup?
is Simple like this: We setup a few Agent Handleron the DMZ, then have some agents( almost 1000) connect to DMZ's AH from internet. Because of Company's Policy, we only allow AH thur Firewall. So can we do agent wakeup from AH or have other way to do it? Hope you get the idea!
I think so
You'll definitely need to allow traffic outbound from the AH on the agent wakeup port, which is 8081 by default: if you can't do that then wakeup calls will not be possible