Load-balancing Workflow Engine

The load-balancing run configuration assigns one entire workflow execution to one Hop Server from a group.

It uses the same server table, algorithms, retry policy, admission control, and assignment ledger as the load-balancing pipeline engine.

Each Hop Server in the group still needs a Hop Server metadata definition.

The engine-specific options sit on the same three tabs as the pipeline engine: Management (poll and export), Load-balancing (run configuration, algorithm, retry, state folder), and Hop Servers.

The run configuration to use on the selected server is usually a local workflow engine on that server. Do not point it back at this load-balancing configuration.

Max concurrent on each server counts unfinished pipelines and workflows, so a mixed pool shares one cap.