Techniques for coordination of application components deployed on distributed virtual machines

Abstract

One embodiment is a method and includes monitoring by a module associated with a first application component installed on a first virtual machine (“VM”) a state of at least one second application component installed on a second VM and on which a state of the first application component is at least partially dependent, in which the state of the at least one second application component is made available by a module associated with the at least one application component; determining that the state of the at least one second application component has changed from a first state to a second state; and updating the state of the first application component based on a current state of the first application component and the second state of the at least one second application component.

Description

TECHNICAL FIELD

This disclosure relates in general to the field of communications networks and, more particularly, to a technique for coordination of application components deployed on distributed virtual machine (“VMs”) in such communications networks.

BACKGROUND

In a system implemented using a collection of non-heterogeneous virtual machines, there are often dependencies between the various VMs, which means that there should be a mechanism by which the VMs can manage their startup ordering such that dependencies are satisfied and each VM has what it needs to operate successfully. When the VMs scale by launching parallel identical instances of a given VM, those instances may be considered to be delivering a “grouped function” within the larger application framework. For example, there are cases in which a certain minimum number of such identical instances must be presented for the “grouped function” to actually be able to provide the shared service. Accordingly, a mechanism for allowing a “derived group state” to be created and shared would be useful in these instances.

Additionally, in connection with an application that includes functional components, or services, distributed across discrete VMs, there is currently no existing framework that aids in the coordination between functional components so that they may be controlled independently of the underlying VMs on which they run. Coordination is generally ad hoc in nature and is therefore subject to incomplete coverage.

BRIEF DESCRIPTION OF THE DRAWINGS

To provide a more complete understanding of the present disclosure and features and advantages thereof, reference is made to the following description, taken in conjunction with the accompanying figures, wherein like reference numerals represent like parts, in which:

FIG. 1 illustrates a cloud computing services stack in accordance with embodiments herein for implementing a technique for coordination of application components deployed on distributed virtual machine (“VMs”);

FIG. 2 is a simplified block diagram of a communication system in which a distributed collection of VMs for running a specific subset of application components is coordinated via an operator using a manual installation process;

FIG. 3 is a simplified block diagram of a communication system in which techniques for coordination of application components deployed on distributed virtual machine (“VMs”) in accordance with embodiments described herein may be implemented;

FIG. 4 is a simplified block diagram of example architecture of an Agent for use in a communication system in which techniques for coordination of application components deployed on distributed virtual machine (“VMs”) in accordance with embodiments described herein may be implemented;

FIG. 5 is a state diagram of a plugin lifecycle in accordance with embodiments for implementing techniques for coordination of application components deployed on distributed VMs in accordance with embodiments described herein;

FIG. 6A is a simplified block diagram of an example shared data store for implementing techniques for coordination of application components deployed on distributed VMs in accordance with embodiments described herein;

FIG. 6B is a flowchart of operational steps that may be performed by a plugin, either alone or in combination with an agent, for implementing techniques for coordination of application components deployed on distributed VMs in accordance with embodiments described herein; and

FIG. 7 is a simplified block diagram of a machine comprising an element of a communications network in which techniques for coordination of application components deployed on distributed VMs in accordance with embodiments described herein may be implemented.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

Overview

One embodiment is a method and includes monitoring by a module associated with a first application component installed on a first virtual machine (“VM”) a state of at least one second application component installed on a second VM and on which a state of the first application component is at least partially dependent, in which the state of the at least one second application component is made available by a module associated with the at least one application component; determining that the state of the at least one second application component has changed from a first state to a second state; and updating the state of the first application component based on a current state of the first application component and the second state of the at least one second application component.

Example Embodiments

Cloud computing may be defined as an Internet-based type of computing that provides shared processing resources and data to computers and other devices on demand. Cloud computing is a model for enabling ubiquitous, on-demand access to a shared pool of configurable resources, including, for example, networks, servers, storage, applications, and services, that may be rapidly provisioned and released for use with minimal management effort. Cloud computing solutions (including cloud storage solutions) provide users with the ability to process and store their data in third-party data centers and rely heavily on resource sharing to achieve coherency and economy of scale. An important benefit of cloud computing is that it allows enterprises to avoid making costly infrastructure expenditures and instead focus funds and efforts on projects that differentiate their business from other businesses. Cloud computing may also enable businesses to make their applications available more quickly and enables IT to more rapidly adjust resources to meet fluctuating demand. The availability of high-capacity networks and low-cost computing and storage devices, as well as the adoption of hardware virtualization, service-oriented computing architectures, and autonomic and utility computing, have resulted in an explosion in cloud computing, which enables enterprises to quickly scale up or down in response to demand.

Cloud computing services may be offered in accordance with a variety of models. FIG. 1 illustrates these various models as a stack 10 including infrastructure-as-a-service (“IaaS”) 12, platform-as-a-service (“PaaS”) 14, and software-as-a-service (“SaaS”) 16. IaaS 12 providers offer computers (typically in the form of VMs) and other resources and abstract the user from the details of the infrastructure. A hypervisor runs the VMs as guests and pools of hypervisors within the cloud operational system can support large numbers of VMs and provide the ability to scale services up or down depending on customer requirements.

PaaS 14 providers may offer a development environment to application developers. PaaS providers may provide customers a computing platform (which may include an operating system), a programming-language execution environment, a database, and a web server. Application developers can develop and run their software solutions on a cloud platform without the cost and complexity of buying and managing the underlying hardware and software layers.

In the SaaS model 16, customers have access to application software and databases, while the SaaS provider manages the infrastructure and platforms that run the applications. In particular, SaaS providers install and operate application software in the cloud and cloud users access the software from cloud clients 18. Cloud users do not manage the cloud infrastructure and platform on which the application runs. SaaS eliminates the need to install and run the application on the cloud user's own computers, which simplifies maintenance and support. Scaling of cloud applications can be achieved by cloning tasks onto multiple VMs at run-time to meet changing work demand. Load balancers may be deployed to distribute work over the set of virtual machines. This process is transparent to the cloud user, who sees only a single access-point.

Referring to FIG. 2, the nature of a system created by utilizing a distributed collection of VMs, represented in FIG. 2 by VMs 20(1)-20(N), each running a specific subset of application components 22(1)-22(N), respectively, creates a situation in which coordination between and among the components must be achieved. As shown in FIG. 2, such coordination may be accomplished by operator intervention through use of an elaborate manual install followed by a startup procedure whereby an operator 24 executes commands and/or scripts to initiate a first functional component (e.g., component 22(1)) on a VM (e.g., VM 20(1)) and monitors its progress before proceeding to initiate other components (e.g., components 22(2)-22(N)), on other VMs (e.g., VMs 20(2)-20(N)) in a manner in which any ordering and dependencies among the components 22(1)-22(N) is satisfied. For cloud-based solutions, particularly where there is a desire to auto-scale, manual intervention is not desirable and is error-prone.

In response, embodiments described herein put into place multiple attributes to create a mechanism by which automation of application component-level coordination may be facilitated. Embodiments described herein may be implemented as an Agent running on each of the VMs and tailored for the specific application components hosted on each VM. The key functions of the Agent include (1) providing a shared space in which to post information (akin to a community bulletin board); (2) providing a mechanism to trigger behavior when information in the shared space is altered; (3) providing a shared lifecycle model so that all distributed components have a mutual understanding of application execution states; (4) providing a plugin mechanism that permits the Agent to be specialized for the application components hosted on the VM; and (5) providing a mechanism for configuration of dependencies between the various application components in the system.

As will be described, embodiments create a fully distributed solution to achieve the goal of application-level coordination. In certain embodiments, there is no “master controller” or “puppet master;” the one common shared entity in the solution is a shared data space that should be a facility that can be made highly available. In certain embodiments, ZooKeeper, available from The Apache Software Foundation of Forest Hill, Md., may be used to implement the shared entity.

Referring now to FIG. 3, illustrated therein is a simplified block diagram of a system 30 for coordination of application components deployed on distributed virtual machine (“VMs”) in accordance with embodiments described herein. As shown in FIG. 3, the system 30 includes a number of VMs, represented in FIG. 3 by VMs 32(1)-32(N), each having an Agent, represented in FIG. 3 by Agents 34(1)-34(N), running thereon. As previously noted, each agent can manage multiple features within the VM on which it is installed to implement various features of embodiments described herein. In particular, as will be described in greater detail hereinbelow each Agent 34(1)-34(N) manages aspects of operation of a set of application components 36(1)-36(N) executing on the respective VM 32(1)-32(N). In accordance with features of embodiments described herein, the agents 34(1)-34(N) communicate with one another and share information via a shared entity, which in the illustrated embodiment is a shared data store 36. The shared data store 36 should provide reasonable distributed capabilities (e.g., locking, some understanding of the assurances provided by the data store in the event it is itself distributed, etc.). As noted above, the shared data store 36 may be implemented using ZooKeeper or another comparable technology, such as Consul, doozerd, etcd, and others.

FIG. 4 illustrates a simplified block diagram of the architecture of an Agent 40. As shown in FIG. 4, the Agent 40 comprises a shared Agent module 41 that can accept one or more plug-ins, represented in FIG. 4 by plugins 42(1)-42(N). Each plugin is created from a common framework, but is customized for the specific application component 44(1)-44(N) being managed. As there is typically a one-to-one correspondence between plugins and application components (or services), reference to a plugin and its corresponding application component/service may be used interchangeably herein. The shared Agent 41 provides a shared environment in which the plugins 42(1)-42(N) operate and also provides utilities to interface with the shared data space 38 (FIG. 3) to log Agent/plugin behavior, to the plugin framework, and to other shared facilities. In certain embodiments, information is logged in the shared data store using facilities provided by the agent and plugin framework. The data is organized in a hierarchical manner so that like-components are aggregated into groups, with each group representing a service or feature implemented within the system. For example, a distributed database such as MongoDB might be deployed such that for the “MongoDB service” to be fully functional, three VM instances need to be running MongoDB. If those VM instances are named “mongo-01,” “mongo-02” and “mongo-03,” those shared data records, in the shared data store, might all be organized under a group named “mongo.” A critical part of the plugin framework is the plugin lifecycle. The goal is to provide a share lifecycle abstraction that all application plugins honor so that the semantic is shared throughout the larger distributed application.

FIG. 5 is a state diagram illustrating a plugin lifecycle 50 in accordance with one embodiment. Referring to FIG. 5, an OFF state 52 is the initial state of a plugin and the state out of which the plugin immediately transitions upon being loaded and initialized by the Agent. From the OFF state 52, the plugin transitions to an IDLE state 54 if any external dependencies of the application component being managed by the plugin are not met and to a STARTING state 56 if such external dependencies have been met. The IDLE state 54 is the state in which the plugin waits if it has any external dependencies that have not been met; once those dependencies are met (or are “available”), the plugin transitions to the STARTING state 56.

The STARTING state 56 is a transitional state in which the plugin stays while it performs all tasks needed to bring the application component to be managed by the plugin into a functional state. Once the application component has been successfully rendered functional, the plugin transitions to a READY state 58. The plugin remains in the READY state 58 as long as the application component managed by the plugin is up, running, healthy, and able to perform its tasks within the system. The plugin exits the READY state 58 and transitions to a STOPPING state 60 if at any point the health of the application component being managed by the plugin fails or if an external dependency exits the READY state. The STOPPING state 60 is a transitional state in which the plugin is performing all tasks needed to bring the application component being managed into a shutdown state. Once the application component is shutdown, the plugin transitions to the IDLE state 54.

This shared understanding of states between all distributed components is important to the overall distributed coordination strategy. A key aspect is whether an application component (referred to for the sake of example as Component A) is READY (i.e., able to perform its tasks within the system) or not. If Component A is not READY, other components that depend on Component A (e.g., Component B and Component C) must wait for Component A to become READY. In other words, until Component A is READY, the external dependency of Component B and Component C on Component A is not met. When Component A is READY, the external dependencies of Component B and Component C are met. By correctly specifying the complete set of dependencies between the application components, it is possible that the VMs can be launched in any order and the applications themselves sort out the order in which the application components are started (via the plugins). Additionally, embodiments described herein provide a framework by which application components may respond to changes within the shared application eco system. As application components come and go, not only can the managing of startup/shutdown order be facilitated, but use of the shared data in the system allows application component details (i.e., IP addresses, whether a specific instance is the “leader” of a group of instances, or any other meaningful details) to adjust their own configurations to reflect those changes within the larger system. In this manner, configuration and reconfiguration can be managed in a distributed manner within this system as well.

Additionally, as will be described in greater detail below, grouping the data within shared data store 38 (FIG. 3) in a hierarchical manner allows the system to consider aggregated collections of like-type application component instances into groups with their own group state. It then becomes possible for a given application component to have dependencies upon an entire group of instances. In turn, that grouping of instances may have their own concept of an aggregate state of readiness or non-readiness. This allows for even more sophisticated dependencies to be expressed by the system.

As noted above, coordinating startup may be facilitated by the creation of a shared application lifecycle and a state machine to implement that lifecycle (e.g., using a plugin). In having a shared lifecycle, all of the various distributed components will have a shared semantic and can use the state of other components in the system to determine their own state transitions. A simple lifecycle can be leveraged to move applications through the basic states of operation. In certain embodiments, distributed component state is made available in a common entity, such as a shared data store (e.g., shared data store 38 (FIG. 3)), which is visible to all components in the system. It will be assumed that the common entity has reasonable distributed processing behaviors and support. A lifecycle monitoring feature implemented by a plugin watches the shared data store for state information updates.

As previously noted, there are a number of actions or operations shared between the various distributed components in connection with the lifecycle. One such action is watching for any dependencies to become available (i.e., move to the READY state 48). An application component moves from the IDLE state 44 to the STARTING state 46 when the component has no dependencies or all the dependencies it does have are also in the READY state. As noted above, the STARTING state 46 is the state in which all operations needed to get the application component functional are performed. Once those operations complete successfully, the component is started and the lifecycle transitions to the READY state 48. This serves as a signal to other distributed components in the system that might have a dependency upon this component that it is ready to perform its specific functions in support of the overall distributed product. This completes the startup scenario.

Like startup, shutdown can have ordering as well. Upon a failure of a dependency or the detection that an application component within the VM has failed in some manner that prevents it from performing its function within the larger system, the lifecycle transitions into the STOPPING state 50. In the STOPPING state 50 all operations needed to bring the distributed application component into a stopped state will be performed. This should leave this particular distributed application component ready to be restarted if its dependencies are still/once again satisfied. Transitioning out of the READY state 48 will also serve as a signal to other portions of the system that may have dependencies upon the distributed component that is in the STOPPING state 50 that it is no longer available to perform its function within the context of the larger system. Once the shutdown operations performed in the STOPPING state 50 have completed successfully, the lifecycle transitions back into the IDLE state 44, watching for dependencies to become READY again. Once they are all READY, the lifecycle-monitoring system will once again transition back into the STARTING state 46, unless a shutdown flag has been set, at which time it transitions back into the OFF state 42 and terminates execution. This will serve to effectively terminate the operations of the functions within the VMs.

FIG. 6A is a simplified block diagram of an example shared data store, designated in FIG. 6A by a reference numeral 60, for implementing techniques for coordination of application components deployed on distributed VMs in accordance with embodiments described herein. For purposes of example herein, it will be assumed that the shared data store 60 is implemented in accordance with the aforementioned ZooKeeper architecture.

As shown in FIG. 6A, for purposes of embodiments described herein, two different categories of records are stored within the shared data store 60, including server records, collectively designated by a reference numeral 62, and services (or components) records, collectively designated by a reference numeral 64. Each of a plurality of server records, represented in FIG. 6A by individual server records 66(1)-66(N), corresponds to a specific VM instance. Services records 64 may include group records, such as group records 68 and 70, and individual service records 72(1)-72(3), which are aggregated under the group represented by group record 68, and individual service records 74(1)-74(2), which are aggregated under the group represented by group record 70. Each of the records is a small JSON structured data snippet. A given server record within the system (e.g., server record 66(1)) might appear as follows:

• • {“cluster”:“cluster01”, “hostname”:“appvm01”, “ip”:“10.225.222.192”, “started”:“2016/05/25 13:32:21 UTC”, “uuid”:“ca37d005-fc16-4517-80d3-de2938a25a39”}The server record gives some very basic information about that VM instance (i.e., appnvm-01) with which it uniquely identifies, and is also a place where over time additional data which might correspond to the VM instance may be shared.

A service record is very similar in nature to the server record, but corresponds to a single service instance (or components) running on a single VM within the system. An example server record (e.g., corresponding to server record 72(1)) might appear as follows:

• • {“cluster”:“cluster-01”, “ip”:“appvm-01”, “name”:“mongo-01”, “state”:“IDLE”, “uuid”:“ca37d005-fc16-4517-80d3-de2938a25a39”, “group”:false}This identifies this as a record for the “mongo-01” service, it is IDLE waiting for its dependencies to be satisfied, and this is an instance rather than group record. This brings us to the issue of “service groups”. In the hierarchy illustrated in FIG. 6A, there are two service group entries, or records, one for a service group designated “mongo” (service group record 68) and one for a service group designated “pcrfclient” (service group record 70). Each of those entries has a JSON record as well. An example of a group record (e.g., corresponding to group record 70) might appear as follows:
  • {“cluster”:“cluster-01”, “ip”:“10.225.222.192”, “name”:“pcrfclient”, “state”:“READY”, “uuid”:“ca37d005-fc16-4517-80d3-de2938a25a39”, “group”:true}

There is a common pattern here as well. In this case, the “state” is the “derived group state” and reflects that the “pcrfclient” instances within that group have determined that that group service has enough of its components deployed and healthy so that the group can provide the desired service to the rest of the system.

All plugins reference the data in this hierarchy to determine, independently, what the state of the rest of the system might be. Based on that evaluation (i.e. what has been specified as a given service's dependencies) it moves through its own lifecycle. As it does so it reflects its own state within its record in the shared store, and that in turn may impact decisions made by other service components within the system.

For example, a distributed database such as MongoDB might be deployed such that for the “MongoDB service” to be fully functional, three VM instances need to be running MongoDB. If those VM instances are named “mongo-01,” “mongo-02” and “mongo-03,” those shared data records, in the shared data store, might all be organized under a group named “mongo.”

Referring now to FIG. 6B, illustrated therein is a flowchart of operational steps that may be performed by a plugin, either alone or in combination with an agent, for implementing techniques for coordination of application components deployed on distributed VMs in accordance with embodiments described herein. In step 80, the plugin monitors the system. This step may be accomplished by the plugin monitoring records in the shared data store corresponding to services/components on which the current plugin has dependencies. Additionally, this step may include monitoring the health and performance of the component (or service) to which the current plugin corresponds. In step 82, a determination is made whether there has been a change in state of any of the plugins on which the current plugin depends (e.g., as indicated in the corresponding individual service records corresponding to those plugins). If a negative determination is made in step 82, execution proceeds to step 84. In step 84, a determination is made whether there has been a change of state, or operational status, of the application to which the current plugin corresponds. For example, the application component may have experienced a failure and is no longer functional. If a negative determination is made in step 82, execution returns to step 80.

If a positive determination is made in either step 82 or step 84, execution proceeds to step 86. In step 86, a new state of the plugin is determined based on the changes detected in step 82 or step 84 and the state is updated in the shared space. It will be recognized that this change in state may cause a cascade of state changes within the shared space if either group states or other plugins depend on the state of the current plugin.

Distributed coordination of separately executing components in a large system, expressed as executing VMs, often have dependencies in place among the VMs. It is also the case that certain subsets of VMs, in the aggregate, provide a single monolithic function to other portions of the system. One example of this is a distributed database, such as Cassandra, in which multiple mostly identical instances of the application executing in a VM are launched to provide execution resiliency, creating a more robust and scalable system. In these scenarios, it is often the case that a certain minimum number of these otherwise identical instances must be present for the group of VMs to provide the desired system functionality. For the sake of example, the certain assumptions will be made. First, for specific engineering reasons, it will be assumed that two instances of a given database must be available before the “database function” within the system is legitimately able to perform services. It will be further assumed that, for normal function, a total of three such instances will be started to provide additional robustness. In this group of three, it will be assumed that one instance will be identified as the “seed” or “lead,” and that without that specially identified instance, the group cannot be considered to perform the desired services. Additionally, it will be assumed that at least one of the two non-lead instances must be functioning (but which one is unimportant to the overall “derived group state.”)

To reflect the derived group state in a distributed manner (i.e., without having some overarching manager or monitor function outside the group itself), it must be possible for the constituents of the group to determine the overall group state between themselves. The advantage of this is that it provides a good separation of concerns in the members of the group and only the members of the group care what constitutes “functional.” It also means that the group can change the meaning of “functional” independently of any need to coordinate with another portion of the system. To have this shared calculation of the “derived group state,” several mechanisms should be available to the individual instances. For example, each instance must be able to monitor the state of other instances within its own group. Each instance must be able to recognize the appearance of another member of the system. Each instance must be able to discern its own role (i.e., “seed” or “lead,” etc.) within the system and must be able to gain and release a lock on a “derived group state” value in the shared data store.

When a change occurs within the group (e.g., a new instance becomes READY or an existing instance becomes IDLE or STOPPED), each other instance within the group recalculates the value of the “derived group state.” This calculation is performed based on the desired criteria (e.g., the number of running instances and the existence of a lead instance) and the value is updated in the shared data store. Prior to the start of those calculations, a lock on the derived group state value must be acquired by the agent updating the value so that updates are serialized. The calculation is then performed, the value updated, and the lock released. The calculation should reflect either READY or one of the not ready states, with various intermediate states being unimportant to the broader system.

If a VM shuts down, the derived group state must be updated as the VM transitions out of operation. While this will occur automatically when more than one VM is existent in the group (see above), in the degenerate case of the final VM of a group shutting down, it will need to perform its own update on the way out for the derived group state to be reflected correctly. While this will result in multiple updates, the net result will be an accurate reflection of the group and no single entity will have the responsibility of maintaining the value. Thus, even in a degenerate case of there being only a single instance available in the system, the group value will be accurate.

Finally, the other portions of the distributed system must be allowed to update the derived group state for a group not their own in the event that the group reflects a READY state, but there are no members in that group remaining. This particular corner case can occur when the last participating member of a group is shut down without the distributed management portion of that VM being given a chance to perform a final update of the derived group state. Given that this is a very exceptional situation (clean shutdowns should be the norm), the occurrence should be rare.

In summary, embodiments described herein provide a framework for cloud-based applications that consist of multiple functional components deployed on a collection of VMs distributed throughout a network whereby the lifecycle of the functional components can be coordinated independently of the lifecycle of the VMs so that the functional components can operate cohesively. The framework described herein enables application component coordination for automated deployment and maintenance of the application. An advantage of these embodiments is that it enables the distributed coordination of complex interactions between components of the system. Responsibility is apportioned to the individual components, which are aware of their own needs, including those other portions of the system upon which they are dependent. In this manner, dependencies can be monitored and satisfied and a complex graph of dependencies can be created without the need for a central controlling entity.

Additionally, embodiments described herein create a normalized shared lifecycle for all cooperating VMs within an application comprised of non-heterogeneous VMs. Use of this lifecycle allows all VMs within the application to have a clear understanding of the state of their peers, and whether those peers are in a state in which they can satisfy their role in the application. A clear lifecycle that facilitates identifying which components of the system are ready to satisfy their role makes it significantly easier to craft distributed applications in a complex deployment. Creating a means by which the concept of a READY state is well-defined enables all other cooperating components in the system to know whether their dependencies are satisfied, thus allowing them to move into a READY state of their own.

Still further, embodiments described herein for aggregating the state of multiple similar VMs into a single aggregate state strives to create a homogenous entity that appears monolithic, but is actually made up of multiple VMs. This allows easier determination of whether adequate service resources are up and running so that service availability goals in a distributed environment may be met. In aggregating the state, a peer entity that depends on the service being monitored can watch a single indicator, rather than canvassing all of the separate entities for their individual states.

Turning to FIG. 7, FIG. 7 illustrates a simplified block diagram of an example machine (or apparatus) 130, which in certain embodiments may be a device on which VMs are deployed, that may be implemented in embodiments described herein. The example machine 130 corresponds to network elements and computing devices that may be deployed in a communications network. In particular, FIG. 7 illustrates a block diagram representation of an example form of a machine within which software and hardware cause machine 130 to perform any one or more of the activities or operations discussed herein. As shown in FIG. 7, machine 130 may include a processor 132, a main memory 133, secondary storage 134, a wireless network interface 135, a wired network interface 136, a user interface 137, and a removable media drive 138 including a computer-readable medium 139. A bus 131, such as a system bus and a memory bus, may provide electronic communication between processor 132 and the memory, drives, interfaces, and other components of machine 130.

Processor 132, which may also be referred to as a central processing unit (“CPU”), can include any general or special-purpose processor capable of executing machine readable instructions and performing operations on data as instructed by the machine readable instructions. Main memory 133 may be directly accessible to processor 132 for accessing machine instructions and may be in the form of random access memory (“RAM”) or any type of dynamic storage (e.g., dynamic random access memory (“DRAM”)). Secondary storage 134 can be any non-volatile memory such as a hard disk, which is capable of storing electronic data including executable software files. Externally stored electronic data may be provided to computer 130 through one or more removable media drives 138, which may be configured to receive any type of external media such as compact discs (“CDs”), digital video discs (“DVDs”), flash drives, external hard drives, etc.

Wireless and wired network interfaces 135 and 136 can be provided to enable electronic communication between machine 130 and other machines, or nodes. In one example, wireless network interface 135 could include a wireless network controller (“WNIC”) with suitable transmitting and receiving components, such as transceivers, for wirelessly communicating within a network. Wired network interface 136 can enable machine 130 to physically connect to a network by a wire line such as an Ethernet cable. Both wireless and wired network interfaces 135 and 136 may be configured to facilitate communications using suitable communication protocols such as, for example, Internet Protocol Suite (“TCP/IP”). Machine 130 is shown with both wireless and wired network interfaces 135 and 136 for illustrative purposes only. While one or more wireless and hardwire interfaces may be provided in machine 130, or externally connected to machine 130, only one connection option is needed to enable connection of machine 130 to a network.

A user interface 137 may be provided in some machines to allow a user to interact with the machine 130. User interface 137 could include a display device such as a graphical display device (e.g., plasma display panel (“PDP”), a liquid crystal display (“LCD”), a cathode ray tube (“CRT”), etc.). In addition, any appropriate input mechanism may also be included such as a keyboard, a touch screen, a mouse, a trackball, voice recognition, touch pad, etc.

Removable media drive 138 represents a drive configured to receive any type of external computer-readable media (e.g., computer-readable medium 139). Instructions embodying the activities or functions described herein may be stored on one or more external computer-readable media. Additionally, such instructions may also, or alternatively, reside at least partially within a memory element (e.g., in main memory 133 or cache memory of processor 132) of machine 130 during execution, or within a non-volatile memory element (e.g., secondary storage 134) of machine 130. Accordingly, other memory elements of machine 130 also constitute computer-readable media. Thus, “computer-readable medium” is meant to include any medium that is capable of storing instructions for execution by machine 130 that cause the machine to perform any one or more of the activities disclosed herein.

Not shown in FIG. 7 is additional hardware that may be suitably coupled to processor 132 and other components in the form of memory management units (“MMU”), additional symmetric multiprocessing (“SMP”) elements, physical memory, peripheral component interconnect (“PCI”) bus and corresponding bridges, small computer system interface (“SCSI”)/integrated drive electronics (“IDE”) elements, etc. Machine 130 may include any additional suitable hardware, software, components, modules, interfaces, or objects that facilitate the operations thereof. This may be inclusive of appropriate algorithms and communication protocols that allow for the effective protection and communication of data. Furthermore, any suitable operating system may also be configured in machine 130 to appropriately manage the operation of the hardware components therein.

The elements, shown and/or described with reference to machine 130, are intended for illustrative purposes and are not meant to imply architectural limitations of machines such as those utilized in accordance with the present disclosure. In addition, each machine may include more or fewer components where appropriate and based on particular needs. As used herein in this Specification, the term “machine” is meant to encompass any computing device or network element such as servers, routers, personal computers, client computers, network appliances, switches, bridges, gateways, processors, load balancers, wireless LAN controllers, firewalls, or any other suitable device, component, element, or object operable to affect or process electronic information in a network environment.

In example implementations, at least some portions of the activities described herein may be implemented in software. In some embodiments, this software could be received or downloaded from a web server, provided on computer-readable media, or configured by a manufacturer of a particular element in order to implement the embodiments described herein. In some embodiments, one or more of these features may be implemented in hardware, provided external to these elements, or consolidated in any appropriate manner to achieve the intended functionality.

In one example implementation, machines on which VMs are deployed may include any suitable hardware, software, components, modules, or objects that facilitate the operations thereof, as well as suitable interfaces for receiving, transmitting, and/or otherwise communicating data or information in a network environment. This may be inclusive of appropriate algorithms and communication protocols that allow for the effective exchange of data or information.

Furthermore, in the embodiments described and illustrated herein, some of the processors and memory elements associated with the various network elements may be removed, or otherwise consolidated such that a single processor and a single memory location are responsible for certain activities. Alternatively, certain processing functions could be separated and separate processors and/or physical machines could implement various functionalities. In a general sense, the arrangements depicted in the FIGURES may be more logical in their representations, whereas a physical architecture may include various permutations, combinations, and/or hybrids of these elements. It is imperative to note that countless possible design configurations can be used to achieve the operational objectives outlined here. Accordingly, the associated infrastructure has a myriad of substitute arrangements, design choices, device possibilities, hardware configurations, software implementations, equipment options, etc.

In some of the example embodiments, one or more memory elements (e.g., main memory 133, secondary storage 134, computer-readable medium 139) can store data used in implementing embodiments described and illustrated herein. This includes at least some of the memory elements being able to store instructions (e.g., software, logic, code, etc.) that are executed to carry out the activities described in this Specification. A processor can execute any type of instructions associated with the data to achieve the operations detailed herein in this Specification. In one example, one or more processors (e.g., processor 132) could transform an element or an article (e.g., data) from one state or thing to another state or thing. In another example, the activities outlined herein may be implemented with fixed logic or programmable logic (e.g., software/computer instructions executed by a processor) and the elements identified herein could be some type of a programmable processor, programmable digital logic (e.g., a field programmable gate array (“FPGA”), an erasable programmable read only memory (“EPROM”), an electrically erasable programmable read only memory (“EEPROM”), an ASIC that includes digital logic, software, code, electronic instructions, flash memory, optical disks, CD-ROMs, DVD ROMs, magnetic or optical cards, other types of machine-readable mediums suitable for storing electronic instructions, or any suitable combination thereof.

Components of communications network described herein may keep information in any suitable type of memory (e.g., random access memory (“RAM”), read-only memory (“ROM”), erasable programmable ROM (“EPROM”), electrically erasable programmable ROM (“EEPROM”), etc.), software, hardware, or in any other suitable component, device, element, or object where appropriate and based on particular needs. Any of the memory items discussed herein should be construed as being encompassed within the broad term “memory element.” The information being read, used, tracked, sent, transmitted, communicated, or received by network environment, could be provided in any database, register, queue, table, cache, control list, or other storage structure, all of which can be referenced at any suitable timeframe. Any such storage options may be included within the broad term “memory element” as used herein. Similarly, any of the potential processing elements and modules described in this Specification should be construed as being encompassed within the broad term “processor.”

Note that with the example provided above, as well as numerous other examples provided herein, interaction may be described in terms of two, three, or four network elements. However, this has been done for purposes of clarity and example only. In certain cases, it may be easier to describe one or more of the functionalities of a given set of flows by only referencing a limited number of network elements. It should be appreciated that topologies illustrated in and described with reference to the accompanying FIGURES (and their teachings) are readily scalable and can accommodate a large number of components, as well as more complicated/sophisticated arrangements and configurations. Accordingly, the examples provided should not limit the scope or inhibit the broad teachings of the illustrated topologies as potentially applied to myriad other architectures.

It is also important to note that the steps in the preceding flow diagrams illustrate only some of the possible signaling scenarios and patterns that may be executed by, or within, communication systems shown in the FIGURES. Some of these steps may be deleted or removed where appropriate, or these steps may be modified or changed considerably without departing from the scope of the present disclosure. In addition, a number of these operations have been described as being executed concurrently with, or in parallel to, one or more additional operations. However, the timing of these operations may be altered considerably. The preceding operational flows have been offered for purposes of example and discussion. Substantial flexibility is provided by communication systems shown in the FIGURES in that any suitable arrangements, chronologies, configurations, and timing mechanisms may be provided without departing from the teachings of the present disclosure.

Although the present disclosure has been described in detail with reference to particular arrangements and configurations, these example configurations and arrangements may be changed significantly without departing from the scope of the present disclosure. For example, although the present disclosure has been described with reference to particular communication exchanges, embodiments described herein may be applicable to other architectures.

Numerous other changes, substitutions, variations, alterations, and modifications may be ascertained to one skilled in the art and it is intended that the present disclosure encompass all such changes, substitutions, variations, alterations, and modifications as falling within the scope of the appended claims. In order to assist the United States Patent and Trademark Office (USPTO) and, additionally, any readers of any patent issued on this application in interpreting the claims appended hereto, Applicant wishes to note that the Applicant: (a) does not intend any of the appended claims to invoke paragraph six (6) of 35 U.S.C. section 142 as it exists on the date of the filing hereof unless the words “means for” or “step for” are specifically used in the particular claims; and (b) does not intend, by any statement in the specification, to limit this disclosure in any way that is not otherwise reflected in the appended claims.

Claims

1. A method for managing an application having functional components distributed on different virtual machines (“VM”) such that each of the virtual machines performs a subset of the application, each virtual machine having an agent that locally controls the functional components of its virtual machine, the method comprising: maintaining a shared data store accessible by the agents, the shared data store including a status of the functional components of the application;monitoring by a first agent associated with a first application component installed on a first VM the shared data store for a state of at least one second application component installed on a second VM and on which a state of the first application component is at least partially dependent;determining by the first agent from the shared data store that the state of the at least one second application component has changed from a first state to a second state;updating by the first agent the state of the first application component based on a current state of the first application component and the second state of the at least one second application component; andupdating the shared data store, by the first agent, to reflect the updated state of the first application;wherein the agents collectively control the application in a distributed fashion without a master control.

2. The method of claim 1 further comprising: detecting a change in an operational status of the first application component.

3. The method of claim 2, wherein the current state of the first application component is a STARTING state in which startup tasks necessary to bring the first application component into a functional state are performed and wherein upon completion of the startup tasks, the method further comprises transitioning the state of the first application component to a READY state in which the first application component is fully operational.

4. The method of claim 2, wherein the current state of the first application component is a READY state in which the first application component is fully operational, and either the state of the at least one second application component is the READY state or the first application component becomes non-operational, the method further comprising transitioning the state of the first application component to a STOPPING state, in which shutdown activities are performed in connection with the first application component, and subsequently transitioning the first application component to an IDLE state when shutdown activities of the first application component have been completed.

5. The method of claim 1, wherein the at least one second application component comprises a plurality of second application components, and wherein the state of the at least one second application component comprises a derived group state representing a collective state of the plurality of second application components.

6. The method of claim 1, wherein the current state of the first application component is an OFF state in which the first application is non-functional and the second state of the at least one second application component is a READY state in which the at least one second application is fully operational, and wherein the updating the state of the first application component comprises transitioning the state of the first application component to a STARTING state in which startup tasks necessary to bring the first application component into a functional state are performed.

7. The method of claim 1, wherein the current state of the first application component is an OFF state in which the first application is non-functional and the second state of the at least one second application component is other than a READY state, and wherein the updating the state of the first application component comprises transitioning the state of the first application component to an IDLE state in which the first application component awaits transition of the at least one second application component to the READY state.

8. One or more non-transitory tangible media having encoded thereon logic that includes code for execution and when executed by a processor is operable to perform operations for managing an application having functional components distributed on different virtual machines (“VM”) such that each of the virtual machines performs a subset of the application, each virtual machine having an agent that locally controls the functional components of its virtual machine, the operations comprising: maintaining a shared data store accessible by the agents, the shared data store including a status of the functional components of the applicationmonitoring by a first agent associated with a first application component installed on a first VM the shared data store for a state of at least one second application component installed on a second VM and on which a state of the first application component is at least partially dependent;determining by the first agent from the shared data store that the state of the at least one second application component has changed from a first state to a second state;updating by the first agent the state of the first application component based on a current state of the first application component and the second state of the at least one second application component; andupdating the shared data store, by the first agent, to reflect the updated state of the first application;wherein the agents collectively control the application in a distributed fashion without a master control.

9. The media of claim 8, wherein the operations further comprise: detecting a change in an operational status of the first application component.

10. The media of claim 9, wherein the current state of the first application component is a STARTING state in which startup tasks necessary to bring the first application component into a functional state are performed and wherein upon completion of the startup tasks, the operations further comprising transitioning the state of the first application component to a READY state in which the first application component is fully operational.

11. The media of claim 9, wherein the current state of the first application component is a READY state in which the first application component is fully operational, and either the state of the at least one second application component is the READY state or the first application component becomes non-operational, the operations further comprising transitioning the state of the first application component to a STOPPING state, in which shutdown activities are performed in connection with the first application component, and subsequently transitioning the first application component to an IDLE state when shutdown activities of the first application component have been completed.

12. The media of claim 8, wherein the at least one second application component comprises a plurality of second application components, and wherein the state of the at least one second application component comprises a derived group state representing a collective state of the plurality of second application components.

13. The media of claim 8, wherein the current state of the first application component is an OFF state in which the first application is non-functional and the second state of the at least one second application component is a READY state in which the at least one second application is fully operational, and wherein the updating the state of the first application component comprises transitioning the state of the first application component to a STARTING state in which startup tasks necessary to bring the first application component into a functional state are performed.

14. The media of claim 8, wherein the current state of the first application component is an OFF state in which the first application is non-functional and the second state of the at least one second application component is other than a READY state, and wherein the updating the state of the first application component comprises transitioning the state of the first application component to an IDLE state in which the first application component awaits transition of the at least one second application component to the READY state.

15. An apparatus comprising: a memory element configured to store data;a processor operable to execute instructions associated with the data; anda plug-in module associated with a first application component installed on a first VM configured to: monitor a shared data store for a state of at least one second application component installed on a second VM and on which a state of the first application component is at least partially dependent;determine from the shared data store that the state of the at least one second application component has changed from a first state to a second state; andupdate the state of the first application component based on a current state of the first application component and the second state of the at least one second application component; andupdating the shared data store, to reflect the updated state of the first application;wherein the module locally controls the first application component of the application in a distributed fashion without a master control.

16. The apparatus of claim 15, wherein the module is further configured to: detect a change in an operational status of the first application component.

17. The apparatus of claim 15, wherein the at least one second application component comprises a plurality of second application components, and wherein the state of the at least one second application component comprises a derived group state representing a collective state of the plurality of second application components.