Looking at the source I can't imagine a reason why the clients list is restricted to only fqdn that had a scrape connection (/poll) in less that the registrationTimeout interval ?
I really don't understand because the underlying channel between clients and proxy are still maintened and the client can still be reached by a scrapping command ( GET http://client:port/ ) even if it has disappeared from the list.
Moreover, to be able to use the /clients url as inventory this means that it is absolutely necessary to connect within the time interval (registrationTimeout) after stopping/restarting the proxy to have the list of all clients.
This is something historicall ?
I was planning to use this url to monitor the connection between proxy and clients (sometimes FW empties sessions, or something else is wrong), but it is not possible because host should disappear anyway!
Looking at the source I can't imagine a reason why the clients list is restricted to only fqdn that had a scrape connection (/poll) in less that the registrationTimeout interval ?
I really don't understand because the underlying channel between clients and proxy are still maintened and the client can still be reached by a scrapping command ( GET http://client:port/ ) even if it has disappeared from the list.
Moreover, to be able to use the /clients url as inventory this means that it is absolutely necessary to connect within the time interval (registrationTimeout) after stopping/restarting the proxy to have the list of all clients.
This is something historicall ?
I was planning to use this url to monitor the connection between proxy and clients (sometimes FW empties sessions, or something else is wrong), but it is not possible because host should disappear anyway!
So if someone has an idea ?