Virtual machine (VM) realm integration and management
A hardware and operating system independent platform and/or run time architecture that implements a configuration and algorithmic based application and process for enabling multiple virtual machines or internal networked physical servers (hosts) to be organized into groups or “realms” in real time and against specific configurable metrics. The configurable metrics include one or more of the following; load or processing capacity availability, cost of CPU processing, location based on risk, importance to business continuity priority requirements, disaster recovery strength and priority, geographic location of host aligned to IP, and any other client specific or custom requirements. A realm manager balances all of these configurable metrics and creates an algorithm to address allocation. The algorithm is fully elastic, allowing for automatic opening and closing of virtual machines to accommodate any internal or external changes.
1. A non-transitory computer readable medium including computer program instructions to cause a computing device to perform steps in a process, comprising:
defining a realm of at least one application cluster group that includes a worker that encapsulates an application function capable of:
performing a processing task as part of a defined workflow of processing tasks and that spans at least one virtual machine hosted on one or more physical nodes, wherein the composition of the application cluster group is defined by the processing requirements of the application function;
managing, by a realm manager of the application cluster group and based on a workload configuration policy, a workload level of each virtual machine;
adjusting a number of virtual machines hosted on the one or more physical nodes in response to a request from the realm manager and based on at least one of monitoring information that pertains to the at least one virtual machine and that is obtained by a virtual machine monitor and diagnostic information that pertains to the at least one virtual machine and that is obtained by a virtual machine monitor;
allocating work, by a distribution manager, to adjust a workload of each application cluster group within the realm;
managing, by a network configurator and in response to information received from the realm manager, a processing capacity load of each application cluster group within the realm;
adjusting a number of workers within the application cluster group in accordance with processing needs determined by a cluster manager of the application cluster group;
receiving, at a work queue, a unit of work from a client application;
managing the work queue with a queue manager of the application cluster group and in accordance with a queue configuration policy;
sending, by a current master dispatcher of the application cluster group, the unit of work to the worker, the current master dispatcher being negotiated from a plurality of slave dispatchers;
performing the processing task encapsulated by the worker on the unit of work in accordance with a defined workflow;
returning a fully processed unit of work to the client application; and
dynamically reforming the application cluster when the current master dispatcher becomes inoperative by the plurality of slave dispatchers negotiating amongst themselves to select one of the slave dispatchers to act as the new master dispatcher.
2. The non-transitory computer readable medium of claim 1 wherein the application cluster group includes two or more workers, each of which encapsulates an application function, and all of which are arranged with respect to each other to implement the defined workflow of processing tasks.
3. The non-transitory computer readable medium of claim 2 wherein the application cluster group is configured to perform parallel processing tasks.