International Business Machines Corporation

Alex Chen

Theodore Timothy Harris, Jr

Angqin Bai

James Chien-Chiung Chen

Minh-Ngoc Le Huynh

Gail Andrea Spear

Carol Santich Mellgren

Olympia Gluck

James C Chen

William Henry Travis

Nicholas D Fifer

Gabriel G Walder

Frederick James Carberry, Ii

Edwin K Scott

Provided is a method for managing operation requests using different resources. In one embodiment, a first queue is provided for operations which utilize a first resource of a first and second resource. A second queue is provided for operations which utilize the second resource. An operation is queued on the first queue until the first resource is acquired. The first resource is released if the second resource is not also acquired. The operation is queued on the second queue when the first resource is acquired but the second resource is not. In addition, the first resource is released until the operation acquires both the first resource and the second resource.

Managing operation requests using different resources

Provided are a system and program for managing operation requests using different resources. In one embodiment, a first queue is provided for operations which utilize a first resource of a first and second resource. A second queue is provided for operations which utilize the second resource. An operation is queued on the first queue until the first resource is acquired. The first resource is released if the second resource is not also acquired. The operation is queued on the second queue when the first resource is acquired but the second resource is not. In addition, the first resource is released until the operation acquires both the first resource and the second resource.

Provided is a method, system and program for managing operation requests using different resources. In one embodiment, a first queue is provided for operations which utilize a first resource of a first and second resource. A second queue is provided for operations which utilize the second resource. An operation is queued on the first queue until the first resource is acquired. The first resource is released if the second resource is not also acquired. The operation is queued on the second queue when the first resource is acquired but the second resource is not. In addition, the first resource is released until the operation acquires both the first resource and the second resource.

Method, system, and program for managing operation requests using different resources

Tracks of data or other data units are copied from a primary storage system to a secondary storage system. For multiple-track or multiple-data unit transfers, a group of tracks or data units which are sequentially related are transferred. A respective task is defined at the primary storage system for monitoring a progress of the transfer of each of the respective data units in a group. A state is maintained for monitoring a progress of the transfer of the group responsive to the progress of each of the tasks. A normal state is selected when there are no failed tasks. A recovery state is selected when there is a failure, and the status of the remaining tasks is being ascertained. A retry state is selected following the recovery state to retry the failed tasks. A failed state indicates an unrecoverable failure.

Method of maintaining task sequence within a task domain across error recovery

An apparatus, system, and method are provided for maintaining task prioritization and load balancing of I/O tasks on servers coupled by a plurality of communication channels. The apparatus includes a task controller for maintaining system resources on a failed target channel while resubmitting an unsuccessful task to a different channel. Also, the apparatus includes an upgrade module for upgrading a task prioritization level of a failed I/O task. Additionally, a load module is provided for balancing I/O tasks evenly across the plurality of communication channels according to a prioritization algorithm. The system includes the task controller and a device controller configured to interface with the task controller. The method includes a computer readable medium with computer readable code for selecting a channel, upgrading the prioritization level of a failed task, and maintaining resources on the failed channel while resubmitting the failed I/O task on a different channel.

Apparatus, system, and method for maintaining task prioritization and load balancing

A method to provide information from a first information storage and retrieval system to a second information storage and retrieval system. Applicants' method provides a first information storage and retrieval system, where that first information storage and retrieval system comprises (N) PPRC adapters and information, and provides a second information storage and retrieval system. The method generates and saves an Established Path Bitmap which recites each of the (N) PPRC adapters, and generates an Available Path Bitmap. If a PPRC adapter is in communication with the secondary information storage and retrieval system, then the method adds that adapter to the Available Path Bitmap. The Established Path Bitmap and the Available Path Bitmap are used for subsequent PPRC operations.

Apparatus and method to provide information from a first information storage and retrieval system to a second information storage and retrieval system

A processor is used to evaluate information regarding the number, size, and priority level of data transfer requests sent to a plurality of communication ports. Additional information regarding the number, size, and priority level of data requests received by the communication ports from this and other processors is evaluated as well. This information is applied to a control algorithm that, in turn, determines which of the communication ports will receive subsequent data transfer requests. The behavior of the control algorithm varies based on the current utilization rate of communication port bandwidths, the size of data transfer requests, and the priority level of the these transfer requests.

Method of balancing work load with prioritized tasks across a multitude of communication ports

Provided is a method, system and program for synchronizing sequential subtasks of a task, so that the subtasks can be dispatched to generate operations in a particular sequence. In one embodiment, a subtask which has been prepared for dispatching is queued in a queue until the queued subtask is the next subtask in sequence for dispatching. In one embodiment, the subtasks may be sorted in the queue by task and by subtask sequence.

Growing community of inventors

Tucson, AZ, United States of America

Alex Chen