IIOP and SOAP Failover in Static Clusters

摘要

Requirement: Clients should experience high-availability when accessing network services. Availability should be transparent and not require altering programs.rnProblem: Availability needs to work on multiple platforms and must not require additional hardware/software, other than the remoting system.rnSolution: Replicate the service in a cluster. Advertise all replica addresses. On client invocation, if a replica address fails, have the client remoting infrastructure try another address.rnAssumption: The number of nodes in the cluster remains constant. Nodes may go up and down, but their number and addresses may not change.rnIIOP Example: Add all cluster instance addresses to IORs. When invoking an IOR try another address if one fails. Cache successful addresses to avoid fallback (important for statefull-session beans).rnExperience: For IIOP communication in an application server the mechanism did not degrade system performance. We have also prototyped this mechanism to work with the JAX-RPC SOAP communications stack.rnConclusion: This failover mechanism provides high-availability for IIOP (and SOAP) communication without requiring additional hardware, software nor alterations to client programs.