Method for managing updates to a distributed network independent of hardware

A system for automatically adjusting a number of application instances in a distributed software-defined network, comprising:

• a plurality of network elements, each network element including a processor and a memory storing instructions that, when executed by the processor, cause the network element to execute one or more fxDeviceApp or fxCloudApp components of a distributed application;
• and
• a distributed resource service (dRS) comprising a processor and a memory storing instructions that, when executed by the processor, cause the dRS to:
• monitor a plurality of distinct usage metrics including CPU utilization and at least one of throughput or latency associated with resource consumption by the fxDeviceApp or fxCloudApp components;
• determine, for each monitored metric, whether a policy-defined threshold condition is satisfied and compute a respective scale recommendation;
• select, in accordance with a policy configured in an application manager and operative to prevent under-provisioning, a resulting desired number of execution environments based on the respective scale recommendations;
• initiate, in response to the resulting desired number exceeding a current number, deployment of additional execution environments for the fxDeviceApp or fxCloudApp components and register the additional execution environments as available service endpoints;
• establish secure communication channels between the additional execution environments and other components of the distributed application using a virtual messaging fabric;
• and
• update, under control of a switch controller, a flow information base of a virtual switch and associated policy rules to include the registered endpoints so that traffic is assigned to the additional execution environments.

A non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a system for automatically adjusting the number of application instances in a distributed software-defined network, cause the system to perform operations including:

• executing, on a plurality of network elements, one or more fxDeviceApp or fxCloudApp components associated with a distributed application;
• monitoring, by a distributed resource service, a plurality of distinct usage metrics including CPU utilization and at least one of throughput or latency;
• for each monitored metric, determining whether a policy-defined threshold condition is satisfied and computing a respective scale recommendation;
• selecting, in accordance with a policy configured in an application manager and operative to prevent under-provisioning, a resulting desired number of execution environments based on the scale recommendations;
• in response to the resulting desired number exceeding a current number, initiating deployment of additional execution environments for the fxDeviceApp or fxCloudApp components and registering the additional execution environments as available service endpoints;
• establishing secure communication channels between the additional execution environments and other components of the distributed application using a virtual messaging fabric;
• and
• updating, under control of a switch controller, a flow information base of a virtual switch and associated policy rules to include the registered endpoints so that traffic is assigned to the additional execution environments.