| 摘要 |
A hive of computing engines, typically including request handlers and process handlers, is used to process information. One implementation includes a request region including multiple request handlers and multiple processing regions, each typically including multiple process handlers. Each request handler is configured to respond to a client service request of a processing job, and if identified to handle the processing job: to query one or more of the processing regions to identify and assign a particular process handler to service the processing job, and to receive a processing result from the particular process handler. Each of the process handlers is configured to respond to such a query, and if identified as the particular process handler: to service the processing job, to process the processing job, to update said identified request handler with state information pertaining to partial processing of said processing job, and to communicate the processing result to the identified request handler.
|
