摘要:
A policy for a computer system is transformed into a plurality of sub-policies, at least one synchronization step, and a final action. The sub-policies are distributed to corresponding sub-components in the computer system. The sub-policies are evaluated, wherein each sub-component evaluates its corresponding one or more sub-policies as other sub-components evaluate their corresponding one or more sub-policies. The sub-policies are synchronized by using the at least one synchronization step. The final action is performed in response to synchronizing the sub-policies. The computer system can be singular or distributed.
摘要:
Formulating an integrated disaster recovery (DR) plan based upon a plurality of DR requirements for an application by receiving a first set of inputs identifying one or more entity types for which the plan is to be formulated, such as an enterprise, one or more sites of the enterprise, the application, or a particular data type for the application. At least one data container representing a subset of data for an application is identified. A second set of inputs is received identifying at least one disaster type for which the plan is to be formulated. A third set of inputs is received identifying a DR requirement for the application as a category of DR Quality of Service (QoS) class to be applied to the disaster type. A composition model is generated specifying one or more respective DR QoS parameters as a function of a corresponding set of one or more QoS parameters representative of a replication technology solution. The replication technology solution encompasses a plurality of storage stack levels. A solution template library is generated for mapping the application to each of a plurality of candidate replication technology solutions. The template library is used to select a DR plan in the form of a replication technology solution for the application.
摘要:
Provided are methods, apparatus arid computer programs for scheduling storage input and/or output (I/O) requests. A method for scheduling storage access requests determines a request processing sequence calculated to maximize SLA-based revenues achievable from processing a number of requests. A storage controller includes a scheduler which implements a revenue-based scheduling function to determine a revenue-maximizing processing sequence, and then assigns storage access requests to locations in a queue corresponding to the determined sequence. In an on-line mode, the scheduler can adapt to additional received requests, evaluating the revenue function for the additional requests and modifying the schedule if required. The method may include analyzing a request stream to predict requests that are likely to be received in the near future, and taking account of the predicted requests when determining a processing schedule.
摘要:
Formulating an integrated disaster recovery (DR) plan based upon a plurality of DR requirements for an application by receiving a first set of inputs identifying one or more entity types for which the plan is to be formulated, such as an enterprise, one or more sites of the enterprise, the application, or a particular data type for the application. At least one data container representing a subset of data for an application is identified. A second set of inputs is received identifying at least one disaster type for which the plan is to be formulated. A third set of inputs is received identifying a DR requirement for the application as a category of DR Quality of Service (QoS) class to be applied to the disaster type. A composition model is generated specifying one or more respective DR QoS parameters as a function of a corresponding set of one or more QoS parameters representative of a replication technology solution. The replication technology solution encompasses a plurality of storage stack levels. A solution template library is generated for mapping the application to each of a plurality of candidate replication technology solutions. The template library is used to select a DR plan in the form of a replication technology solution for the application.
摘要:
Provided are methods, apparatus arid computer programs for scheduling storage input and/or output (I/O) requests. A method for scheduling storage access requests determines a request processing sequence calculated to maximize SLA-based revenues achievable from processing a number of requests. A storage controller includes a scheduler which implements a revenue-based scheduling function to determine a revenue-maximizing processing sequence, and then assigns storage access requests to locations in a queue corresponding to the determined sequence. In an on-line mode, the scheduler can adapt to additional received requests, evaluating the revenue function for the additional requests and modifying the schedule if required. The method may include analysing a request stream to predict requests that are likely to be received in the near future, and taking account of the predicted requests when determining a processing schedule.
摘要:
A policy for a computer system is transformed into a plurality of sub-policies, at least one synchronization step, and a final action. The sub-policies are distributed to corresponding sub-components in the computer system. The sub-policies are evaluated, wherein each sub-component evaluates its corresponding one or more sub-policies as other sub-components evaluate their corresponding one or more sub-policies. The sub-policies are synchronized by using the at least one synchronization step. The final action is performed in response to synchronizing the sub-policies. The computer system can be singular or distributed.
摘要:
A virtual machine placement framework is described to enable a data center operator to develop a placement scheme to satisfy its particular constraints while simultaneously optimizing resource utilization. To generate a placement solution, the virtual machine placement problem is first characterized as a “bin packing” problem. The framework provides simple interface tools and processing modules, and a pluggable architecture for receiving placement algorithms. To generate a solution, an administrator creates an XML representation that abstracts physical entities (e.g., data center, subnet, rack, physical server, and the like) into a hierarchical tree of bins. The administrator also defines a set of “rules” that govern (direct) the placement of the virtual machines by placing constraints on the placement scheme. Using the hierarchical tree and the rules, the framework is executed to generate a placement as a solution to a bin packing problem, preferably on a layer-by-layer basis.
摘要:
A method for scaling is provided. The method of scaling may include determining a predicted workload of a multi-tier application for a defined time period. A change in application capacity for the predicted workload may be determined. A scaling strategy may be dynamically selected for the change in application capacity. The scaling strategy may include vertical scaling and horizontal scaling.
摘要:
Systems determine workload resource usage patterns of a computerized workload, using a computerized device. Such systems use the computerized device to place the computerized workload with a computer server cluster within a private cloud computing environment. Also, systems herein place the computerized workload on a selected computer server within the computer server cluster that has a resource usage pattern complementary to the workload resource usage profile, also using the computerized device. The complementary resource usage pattern peaks at different times from the workload resource usage patterns.
摘要:
Techniques for dynamically selecting a server state for one or more servers in a cluster of servers are provided. The techniques include tracking each active and sleep state of each server in a cluster of servers, and selecting a server state for one or more servers in the cluster of servers to meet one or more workload level requirements of the cluster of servers, wherein selecting a server state for one or more servers comprises scheduling a transition between one or more active and sleep states for the one or more servers, wherein scheduling the transition comprises using power consumption information for each state and transition time information for each transition.