To monitor the status of managed products, MCP agents poll Control Manager based on a schedule. Polling occurs to indicate the status of
the managed product and to check for commands to the managed product from Control Manager. The Control Manager web console then presents the product status. This means that
the managed product’s status is not a real-time, moment-by-moment reflection of the
network’s status. Control Manager checks the status of
each managed product in a sequential manner in the background. Control Manager changes the status of managed products
to offline when a fixed period of time elapses without a heartbeat from the managed
product.
Active heartbeats are not the only means Control Manager determines the status
of managed products. The following also provide Control Manager with the managed
product’s status:
-
Control Manager receives
logs from the managed product. Once Control Manager receives
any type of log from the managed product successfully, this implies
that the managed product is working fine.
-
In two-way communication mode, Control Manager actively
sends out a notification message to trigger the managed product
to retrieve the pending command. If server connects to the managed
product successfully, it also indicates that the product is working
fine and this event counts as a heartbeat.
-
In one-way communication mode, the MCP agent periodically
sends out query commands to Control Manager.
This periodical query behavior works like a heartbeat and is treated
as such by Control Manager.
The MCP heartbeats implement in the following ways:
-
UDP: If the product can reach the server
using UDP, this is the lightest weight, fastest solution available.
However, this does not work in NAT or firewall environments. In
addition, the transmitting client cannot verify that the server
does indeed receive the request.
-
HTTP/HTTPS: To work under a NAT or firewall
environment, a heavyweight HTTP connection can be used to transport
the heartbeat
Control Manager supports both
UDP and HTTP/HTTPS mechanisms to report heartbeats. Control Manager server finds
out which mode the managed product applies during the registration
process. A separate protocol handshake occurs between both parties
to determine the mode.
Aside from simply sending the heartbeat to indicate the product
status, additional data can upload to Control Manager along
with the heartbeat. The data usually contains managed product activity
information to display on the console.
