Hitachi, Ltd.

Hiroyuki Kumazaki

Naoki Utsunomiya

Koji Sonoda

Takashi Nishikado

Akihiro Ito

Kazuhiro Sato

Naonobu Sukegawa

Tetsuhiko Fujii

Kei Nakajima

Kouji Sonoda

Masakazu Fukagawa

Akihiro Nakaya

Toshiya Kurakake

In a system of the present invention in which a file ID is defined for each file, upon receiving a request for accessing a file specifying a file ID from a user, a file server refers to a file management table to obtain a logical disk ID for accessing the file. The file server then refers to a logical disk management table to obtain an I/O path corresponding to the logical disk ID, and accesses a physical disk device by use of the obtained I/O path. When a fault has occurred in an operational I/O path, a file server rewrites the logical disk management tables in all nodes to change the I/O path. This reduces the time taken to switch I/O paths, and hiding the process of switching the I/O paths from the user.

File system for creating switched logical I/O paths for fault recovery

A storage medium for a file input/output control system having a plurality of first computers each having a plurality of disks and connected to a network and at least one second computer connected to the network for accessing the disks connected to the first computers. The storage medium includes a data code section storing a step of retrieving a plurality of first data access requests issued from a plurality of processes of an application and comparing the first data access requests with correspondence relation defining information to confirm that the first data access requests are accesses to the disks, a data code section storing a step of creating a plurality of second data access requests to the disks from the first data access requests confirmed to be accesses to the disks and transmitting the second data access requests to the network, and a data code section storing a step of, in each first computer, rearranging the second data access requests for each disk in the order of block numbers in each disk.

High speed file I/O control system with user set file structure to effect parallel access pattern over a network

In collective I/O processing in which a plurality of processes access the same file in a shared manner, when each user process issues an I/O request, the user process notifies the system of information on a file region accessed by all the processes; information for specifying whether or not a prefetch is performed, and information for specifying whether or not a disk preallocate is performed, together with the I/O requests issued by all associated processes, as hint information. The system provides a buffer for storing data in a file region accessed by all the processes based on the hint information, such that the I/O processing for all the processes is performed using this buffer.

File I/O control method

An illustrative embodiment of the present invention disclosed herein includes a file server having a file management table to obtain a logical disk ID for accessing files. A logical disk management table is consulted to obtain an I/O path corresponding to the logical disk ID, thus accessing the physical device. When a fault occurs along an I/O path, the appropriate tables are altered accordingly to reflect a new I/O path.

A file input/output control system has first computers each having disks and connected to a network; and at least one second computer connected to the network for accessing the disks connected to the plurality of first computers, the second computer having: a retriever for retrieving first data access requests issued from a plurality of processes of an application and comparing the first data access requests with correspondence relation defining information to thereby confirm that the first data access requests are accesses to disks, the correspondence relation defining information being entered by a user in advance and indicating a correspondence relation between the disks and each of regions in a file stored in disks of the first computers; and a scheduler for creating second data access requests to the disks from first data access requests confirmed to be accesses to disks, in accordance with the correspondence relation defining information between the disks and each of the regions in a file stored in the disks of the first computers, and transmitting the second access requests to the network, wherein the first computers access the disks in accordance with the second data access requests received via the network. Each of the first computers has a rearranger for rearranging second data access requests for each of the disks in the order of block numbers in each of the disks. Each of the first computers has a merger for merging, upon detection of that the second data access requests to each of the disks contain data access requests to a continuous disk field, the disk access requests to one disk access request. Although two-dimensional array data distributively stored in the first computers is defined in a row direction, in response to an access command for referring to the two-dimensional array data in a column direction, the network driver reads data containing unnecessary data from the disks, transmits the data containing unnecessary data to each of the second computers, and each of the second computers filters the data containing unnecessary data to discard the unnecessary data and obtain necessary data for the second computer.

High speed file I/O control method and system with user set file

Multiple parallel-job scheduling method and apparatus are provided which can improve the utilization of all processors in a system when a plurality of parallel jobs are executed concurrently. A plurality of processors constituting a computer system and each having the equal function are logically categorized into serial processors for executing a serial computing part or a parallel computing part of a parallel job and a parallel processor group consisting of multiple processors for executing the parallel computing part of the parallel job in parallel. In order that the parallel processors are shared by a plurality of parallel jobs, a synchronization range indicator is provided which can control by program whether the parallel processors are available in correspondence to the respective serial processors. In response to a request for using the parallel processors from a serial processor for which the parallel processors are so set as to be available by means of the synchronization range indicator, operation can be carried out without invoking an interrupt.

Multiple parallel-job scheduling method and apparatus

A method and system for reducing overhead time required to search for memory blocks having programmed executable tasks. The system includes a memory control block queue for managing all memory control blocks, and an executable task control block queue which is a subset of the memory control block queue and which is regularly updated to include memory control blocks having pointers to only programmed currently-executing tasks and programmed executable tasks. When a search for executable tasks is initialized, only the executable task control block queue need be searched. Unnecessary overhead time needed to search memory control blocks having only unexecutable tasks is eliminated.

Growing community of inventors

Yokohama, Japan

Hiroyuki Kumazaki