CddMetricType

• Base type Enumeration

This textual convention defines metric types in the DD. There are many criteria which the DD system can use when sorting a list of IP addresses. In order to collect all of the information used for sorting, the DD system can use one or many of the different 'metrics' it has built in. Once a set of metrics are selected and the information has been gathered, there are many different ways to put this information together and arrive at a sorted list. The currently available metrics are as follows. DRP external metric (drpExternal(1)): This metric is gathered by sending a Director Response Protocol (DRP) request to all DRP associated routers, asking them for the number of BGP 'hops' between them and the client originating the DNS/HTTP query/request. DRP internal metric (drpInternal(2)): This metric is gathered by sending a DRP request to all DRP associated routers, asking them for the distance from themselves to the edge of their BGP AS in the direction of the client originating the DNS/HTTP query/request. DRP server metric (drpServer(3)): This metric is gathered by sending a DRP request to all DRP associated routers, asking them for the distance between themselves and their associated remote server. DRP RTT metric (drpRtt(4)): This metric is gathered by sending a DRP request to all DRP associated routers, asking them for the round-trip time between themselves and the client originating the DNS/HTTP query/request. This metric can be used to direct the client to the server that is likely to have a minimum link latency to the client. Administrative preference metric (admin(5)): This metric can be used in order to specify a simple preference of one server over another. Random metric (random(6)): This metric simply selects a random number for each IP address. If only this metric were used, it would result in selecting an IP address at random from the list. Portion metric (portion(7)): This metric enables users to assign a load portion to each server such that servers with a higher portion value will receive a larger percentage of connections at any one time. The Distributed Director uses these portion metric values to determine the percentage of the current number of requests to give to each server. DFP availability metric (dfpAvailability(8)): This metric enables the Director to make server selection based on server load. The DD obtains the server load information from all Dynamic Feedback Protocol (DFP) agents running on local server load balancers. Servers with less load will receive more connections from clients. Route map metric (routeMap(9)): This metric uses the Cisco route map mechanism to perform server selection based on the route from the servers to the client originating the DNS/HTTP query/request.