摘要:
Systems and methods for providing a development and deployment platform for Software as a Service (SaaS) providers are disclosed. The platform may provide an application programming interface (API) through which customers may configure virtualized grids of language-level virtual machines (VMs), networks, and storage resources. A grid may be horizontally scalable and programmatically configurable by a customer and/or customer code. The platform may allow dynamic scaling of resources (up and down) to match changing demands, by allowing customers and/or customer code to programmatically alter their own system level structure at runtime, through the API. The grid may be accessed by both on-grid and off-grid processes, through internal and external connectivity resources provided by the platform. The platform may provide common utilities for end-user services, deployment operations, and development tasks (e.g., debugging), and may include frameworks and macro-components (e.g. web servers). The platform may host long-running Internet-based services and/or one-shot computational services.
摘要:
Systems and methods for providing a development and deployment platform for Software as a Service (SaaS) providers are disclosed. The platform may provide an application programming interface (API) through which customers may configure virtualized grids of language-level virtual machines (VMs), networks, and storage resources. A grid may be horizontally scalable and programmatically configurable by a customer and/or customer code. The platform may allow dynamic scaling of resources (up and down) to match changing demands, by allowing customers and/or customer code to programmatically alter their own system level structure at runtime, through the API. The grid may be accessed by both on-grid and off-grid processes, through internal and external connectivity resources provided by the platform. The platform may provide common utilities for end-user services, deployment operations, and development tasks (e.g., debugging), and may include frameworks and macro-components (e.g. web servers). The platform may host long-running Internet-based services and/or one-shot computational services.
摘要:
Methods and systems are provided for performing efficient association traversal operations in a Common Information Model (CIM) implemented environment. A CIM Object Manager creates and defines instance objects and corresponding association objects in a repository. The CIM Object Manager creates wrappers for each object that include reverse links that show the relationship from an object to any corresponding association objects. When a relationship request for a selected object is received by the CIM Object Manager, it performs an association traversal operation using the reverse links stored in the repository. The CIM Object Manager locates the target object's wrapper in the repository, and collects relationship information from the reverse links included within the wrapper. The collected relationship information is then passed to the entity that initiated the request. Accordingly, efficient association traversals may be performed, without having to traverse each instance of defined association objects.
摘要:
A transport neutral technique allows an object manager to communicate with a CIM repository using any of a variety of protocols. The object manager software is independent of the transport mechanism used and need not be changed if the transport mechanism changes. A computer system to be managed includes a CIM object manager and any number of provider APIs that provide resource information about the computer system. A CIM repository stores classes and instances used by the object manager. A remote application computer runs a software management application that communicates with the object manager of the computer system using a local client API. A Repository API of the object manager includes an interface definition defining all methods called by the object manager. Also included is a protocol-specific class that implements the interface definition; there is a protocol-specific class for each protocol desired to be supported. Each class implements methods using a specific protocol. A factory class is executed when the object manager invokes a method call passing in a desired protocol parameter. The factory class creates a protocol-specific object of one of the protocol-specific classes depending on the protocol parameter. The object is returned to the object manager which executes one of its protocol-specific methods thus allowing communication to a repository using a protocol independent of the object manager.
摘要:
A replay attack prevention technique uses a secret algorithm exchanged between client and servers computers. The algorithm is used to periodically alter a special replay key. The replay key may then be used to alter the message sent or alter the digest appended to the message such that the message and or its digest varies as time progresses. An unscrupulous third party who replays a stolen operation message would be unaware of the replay key and its variation. Any stolen message that is replayed by the third party would not include the time varying aspect as expected by the server computer. The server computer would thus be able to detect the attempted fraud. The secret replay key may be exchanged during the communication session in an encrypted form, or may be exchanged earlier.