| 摘要 |
Compute resources of multiple resource cards are assigned to compute resource pools. Each compute resource pool is typically associated with a specific service (e.g., VoIP, video service, deep packet inspection, etc). Compute resource groups are created in each compute resource pool and are allocated one or more compute resources of that compute resource pool. Those compute resources in a given resource pool that are not allocated to a compute resource group are set as backup compute resources. Upon a failure of a compute resource in a compute resource pool that includes backup compute resources, a backup compute resource is selected and takes over the function of the failed compute resource. Upon a failure of a compute resource in a compute resource group of a compute resource pool that does not include a backup compute resource, the traffic is load balanced across the remaining compute resources of that compute resource group. |
