| 摘要 |
The present invention relates to distributed networking, and in particularly to a message-based networking environment. More particularly, it relates to business applications, which access the network via some middleware within a predetermined load schedule. In order to alleviate the disadvantages of prior art it is proposed to perform, a method with the following sequence of steps: monitoring, metering the current workload, storing metered workload data in a first appropriate data store, store state information data about the plurality of currently used instances in a second appropriate data store, analyzing both, said stored, metered workload data and said state information for compliance with a predetermined quality-of-service policy, adapting the number of instances usable for satisfying the workload to the workload needs as required by said predetermined quality-of-service policy, wherein meta data of newly created instances are managed in a plurality of message stores, the data contents of which are frequently synchronized in order to decouple the plurality of client applications from bindings fixed during their operation to particular service instances.
|
