Cisco Technology, Inc.

Jacob Hartinger

Burjiz F Pithawala

Narayanan Thyagarajan

Faiyaz Shahpurwala

A method and system for measuring availability of a high-availability network that includes interconnected routers and switches. One embodiment of the present invention provides a network availability monitoring device that is coupled to the network. In this embodiment, ICMP (Internet Control Message Protocol) pings are sent from the network availability monitoring device at regular intervals to the leaf nodes only. Network availability is then determined based on the total number of ICMP pings attempted and the total number of ping responses returned. In this way, when a non-leaf node goes down, the network availability monitoring device may not report this as non-availability if leaf nodes are not affected. One embodiment of the present invention employs a rule-based detection algorithm that detects leaf nodes in the network. Areas of the network that share similar geographic, topology or service characteristics may be grouped together as availability groups. Group availability can be determined based on a total number of ICMP pings sent to the leaf nodes of the availability groups and a total number of ping responses sent by the leaf nodes of the groups. The present invention also provides for an auto-annotation mechanism that assists network engineers in pinpointing root causes of a network problem.

Growing community of inventors

Palo Alto, CA, United States of America

Jacob Hartinger