The model
It helps to separate the model into three kinds of thing:
- Definitions describe applications, reusable configurations and available infrastructure.
- Live resources are productions, runtimes, nodes and migrations that exist right now.
- Read models such as placements, the schedule, alerts and the audit log explain what the Manager did and what it can currently see.
The rest of this section works through them:
- From a product to running work — license, product, product template, production template, production, runtime.
- Capacity and placement — pools, tiers, clouds, machine types, nodes, and how the Manager decides where work runs.
- Keeping work available — cordon, drain, migration, eviction, hot spares.
- Operating and understanding the system — the schedule, alerts, the audit log and operator access.
The API reference gives the exact fields, states and operations for these concepts.