摘要 |
With fast notification of changes to a clustered computing system, through which a number of events are published for system state changes, application s can quickly recover and sessions can quickly be rebalanced. When a resource associated with a service experiences a change in status, such as a termination or a start/restart, a notification event is immediately publishe d. Notification events contain information to enable subscribers to identify, based on matching a session signature, the particular sessions that are affected by the change in status, and to respond accordingly. This allows sessions to be quickly aborted and ongoing processing to be quickly terminat ed when a resource fails, and allows fast rebalancing of work when a resource i s restarted. |