Ensuring high availability of critical services in face of attacks, overload conditions, or network outages is an important requirement in many areas. This project is investigating a system architecture for building highly available services that are resilient to such adverse conditions. The objective is to facilitate construction of adaptive and autonomic services incorporating recovery and protection mechanisms in their deployment to ensure high availability. The techniques being investigated are based on dynamic replication, relocation and regeneration of services in case of overload conditions. Design and development of distributed mechanisms to actively monitor the components of a service in different domains and their operating conditions is an important requirement towards this research goal.

This project is being conducted using the facilities of the PlanetLab infrastructure. This research is utilizing the mobile agent technology for deployment, relocation, and replication ofservice components. In the system architecture being investigated, mobile agents are utilized as mobile service containers to support relocation of a service over the Internet. This project is also investigating an agent-based framework for monitoring the operating conditions of services and their hosting environments at different nodes on the PlanetLab. This research is identifying metrics and models for resource utilization, load conditions, and available capacities at different nodes on the PlanetLab to drive the autonomic mechanisms for service relocation, regeneration, and replication. The PlanetLab environment poses unique challenges as the resource capacities available to a service are based on the proportional share model, and it can change unpredictably due to usage by other users and applications.  For finding location information of mobile and replicated services, this project has developed a DHT based facility using the Pastry system. This DHT based system is used to direct client requests to service replicas based on different kinds of service-access models such as anycast and multicast.

Our current activities on this project have focused on the design and evaluation of a system for deploying highly available and migratable services in shared infrastructures, such as the PlanetLab,
where the available resource capacities at a node can fluctuate significantly. A migratable service can monitor its operating conditions and autonomously relocate itself to another node when the available resource capacities at the current node fall below certain acceptable limits. We investigate here the mechanisms for service relocation, and client-side protocols to access migratory services. The ``blackout periods'', i.e. the time during which the clients are unable to access a migrating service, are needed to be minimized and kept within some tolerable limits for services required to be highly available. We designed and implemented a migratable service using a mobile agent, and evaluated its performance in terms of the blackout periods and the service agent's abilities to autonomously migrate in the network. We utilized replication of service agents to reduce the blackout periods, and developed the coordination protocols for autonomous agent migration in a group of service agents. We also developed an infrastructure service for monitoring the PlanetLab nodes for available resource capacities in order to assist a migratory service in selecting a target node for relocation.
 

Sponsors