Author: Fan Kai Nang(范啟能)
'YAWL for cluster' is an experimental project on making YAWL, an traditional workflow engine, into a distributed engine with minimal modifications.
We extend the engine with an interface, which we called interface C. Like other interfaces, interface C exposes limited but enough engine information and operations, so that an external management application can access and control the engine.
1. Proxy
1. Heartbeat.
Checks if the engines are still available.
2. Workitem scheduling
Switch active workitems to an available engine when the old engine is down.
3. Load balance
Monitor the distribution of engines on hosts, adjust in reference of allocating speed and historical allocating speed.
4. OpenStack adapter
1. Session management (single sign on)
2. Backup cluster management
3. Security