US 8,713,096 B2Granted

State control of remote hosts for management of distributed applications

AssigneeMicrosoft Corporation

InventorsYi Lin, Pavel A. Dournov, Anil Ingle, Frederick J. Smith, IV

Life of this patent

Estimating the term…

Loading patent…

Abstract

Configuring a node. A method includes determining a current state of a target node. The current state of the target node refers to objects currently deployed or running in the remote target node. It includes at least one of an operating system, one or more applications, or configuration setting at the target node. The method further includes determining a desired state for the target node to change the current state. The method further includes accessing a dependency graph based on the version of an agent running on the target node. The dependency graph is supplied to a state machine. The state machine is particular to the target node. The state machine using the current state of the target node, the desired state and the dependency graph, performs operations to drive the target node toward its goal.

Claims23

Dependent claims

A computer-implemented method for dynamically managing a distributed computing environment by changing the state of one or more servers which serve as target nodes in response to service requests from one or more client computer systems which access the distributed computer environment for services, the computer-implemented method comprising:

receiving within a distributed computing environment a request for services from a client computer system;
in response to the request for services, instantiating
a state driver that determines a target node's current state, which comprises at least one of an operating system, one or more applications, or configuration setting at the target node;
and
a goal state builder that determines a new desired state that will be required to meet the request for services, which comprises altering at least one or more of the current state's operating system, one or more applications, or configuration setting;
the state driver then comparing the current state and the new desired state and generating required actions for changing the current state in order to move to the new desired state at the target node, with the required actions being represented as a dependency graph which defines a sequence for carrying out the required actions;
based on the sequence of actions represented in the dependency graph, the state driver causing each required action to be carried out until the new desired state is achieved at the target node;
and
using the new desired state at the target node to respond to the request for services received from the client computing system.

6 more elements

A computer program product comprising one or more physical memory devices which contain executable instructions which, when executed by one or more processors, cause the processors to implementing a method for dynamically managing a distributed computing environment by changing the state of one or more servers which serve as target nodes in response to service requests from one or more client computer systems which access the distributed computer environment for services, the implemented method comprising:

receiving within a distributed computing environment a request for services from a client computer system;
in response to the request for services, instantiating
a state driver that determines a target node's current state, which comprises at least one of an operating system, one or more applications, or configuration setting at the target node;
and
a goal state builder that determines a new desired state that will be required to meet the request for service, which comprises altering at least one or more of the current state's operating system, one or more applications, or configuration setting;
the state driver then comparing the current state and the new desired state and generating required actions for changing the current state in order to move to the new desired state at the target node, with the required actions being represented as a dependency graph which defines a sequence for carrying out the required actions;
based on the sequence of actions represented in the dependency graph, the state driver causing each required action to be carried out until the new desired state is achieved at the target node;
and
using the new desired state at the target node to respond to the request for services received from the client computing system.

6 more elements

A computer system for dynamically changing the state of at least one server which serves as a target node in a distributed computing environment, the system comprising:

one or more processors;
and
one or more physical memory devices which contain executable instructions which, when executed by the one or more processors, cause the processors to implement a method for dynamically changing the state of the target node in response to a service request from a client computer system which accesses the distributed computer environment for services, the implemented method comprising:
in response to the service request, instantiating
a state driver that determines a target node's current state, which comprises at least one of an operating system, one or more applications, or configuration setting at the target node;
and
a goal state builder that determines a new desired state that will be required to meet the service request, which comprises altering at least one or more of the current state's operating system, one or more applications, or configuration setting;
the state driver then comparing the current state and the new desired state and generating required actions for changing the current state in order to move to the new desired state at the target node, with the required actions being represented as a dependency graph which defines a sequence for carrying out the required actions;
based on the sequence of actions represented in the dependency graph, the state driver causing each required action to be carried out until the new desired state is achieved at the target node;
if the state driver detects that the current state and/or the new desired state of the target node have changed before the actions represented by the dependency graph have been fully implemented, then the state driver and/or the goal state builder updating the current state and/or desired new state and generating any new actions required for changing the current state to the new desired state in accordance with how either or both of them have been updated, and then determining which parts of the dependency graph require the updated actions to be performed by the state driver;
and
using the new desired state at the target node to respond to the request for services received from the client computing system.

9 more elements

Dependencies are inferred from the claim text.