High Availability for Network Management Applications

While hardware failures may cripple a network, network service providers are increasingly finding that software or its configuration is a greater source of risk. In today's world of ubiquitous computing there is an everlasting need for interconnecting devices. The continuously growing internet has accelerated this need at a pace like never before. This has resulted in a lot of investment in networking infrastructure such as routers, switches, hubs and the like. This ever-growing infrastructure needs to be manageable in a cost effective manner so as to derive the maximum benefit out of it. For the sake of simplicity we will stick to the traditional definition of management to mean FCAPS. The aim of this paper is to highlight some of the technical challenges in achieving high availability (HA) at the application layer (L4-L7). While the entire paper is about HA for network management applications the principles and guidelines apply equally well for other enterprise applications also. Also as there is a significant industry momentum in the direction of adopting Java and its related technologies as the platform of choice for building manageability solutions, the scope of this paper is restricted to exploring availability considerations for such applications