-
1.
公开(公告)号:WO1994022082A1
公开(公告)日:1994-09-29
申请号:PCT/US1994003188
申请日:1994-03-23
Applicant: ECLIPSE TECHNOLOGIES, INC.
IPC: G06F11/00
CPC classification number: G06F11/201 , G06F2211/103 , G06F2211/1045 , G11C29/88
Abstract: The disk array control system is a fault tolerant controller for arrays of hard disk drives. With the controller (8) as a front end, an array of hard disk drives (12) would appear as a single drive to a host computer system (7) connected to the controller. The controller translates input/output data requests (300) from the host system (7) to input/output data transfer requests for the appropriate drives in the drive array (12). To minimize latency, translation techniques provide for a minimal number of accesses to the drives in the array. Queued input/output requests are supported, and, to maximize throughput, optimal scheduling techniques are used to optimize resource usage and minimize drive access overheads. Means are provided for detecting module failures, sustaining operation after failure, and replacing faulty modules without interrupting service.
Abstract translation: 磁盘阵列控制系统是用于硬盘驱动器阵列的容错控制器。 以控制器(8)作为前端,硬盘驱动器(12)阵列将显示为连接到控制器的主机系统(7)的单个驱动器。 控制器将来自主机系统(7)的输入/输出数据请求(300)转换成对于驱动器阵列(12)中的适当驱动器的输入/输出数据传送请求。 为了最小化延迟,翻译技术提供对阵列中的驱动器的最少访问数量。 支持排队的输入/输出请求,并且为了最大化吞吐量,使用最优调度技术来优化资源使用并最大限度地降低驱动器访问开销。 提供了用于检测模块故障,故障后维持操作以及更换故障模块而不中断服务的方法。