公开(公告)号：US07171516B2

公开(公告)日：2007-01-30

申请号：US10674871

申请日：2003-09-29

申请人： Steven R. Lowe ,  Kevin J. Ash

发明人： Steven R. Lowe ,  Kevin J. Ash

IPC分类号： G06F12/00

CPC分类号： G06F3/0659 ,  G06F3/0613 ,  G06F3/0673

摘要： A data storage control unit is coupled to one or more host devices and to one or more physical storage units, the physical storage units collectives configured as a plurality of logical storage ranks. The storage control unit receives and processes write requests from the host devices and directs that data updates be stored in a temporary storage. The data updates are subsequently destaged to the logical ranks. Write requests are processed at a predetermined rate relative to the rate at which destages are performed (destage mode or rate). The storage control unit evaluates workload conditions of the temporary storage and modifies the destage mode if the temporary storage is in danger of becoming backlogged or if an actual backlog is created, thereby applying a “throttle” to the host write requests. Thus, method, apparatus and program product are provided to dynamically modify a level of throttling whereby through-put of a host device is substantially maintained.

摘要翻译： 数据存储控制单元耦合到一个或多个主机设备和一个或多个物理存储单元，物理存储单元集合被配置为多个逻辑存储器排序。 存储控制单元接收并处理来自主机设备的写入请求并指示数据更新被存储在临时存储器中。 数据更新随后转移到逻辑排名。 写入请求以相对于执行去往的速率的预定速率进行处理（出租模式或速率）。 存储控制单元评估临时存储器的工作负载条件，并且如果临时存储器处于积压的危险中或者如果实际的积压被创建，则修改出租模式，从而对主机写入请求进行“节流”。 因此，提供了方法，装置和程序产品以动态地修改节流水平，从而基本维持主机设备的通过。

公开(公告)号：US20080040553A1

公开(公告)日：2008-02-14

申请号：US11464113

申请日：2006-08-11

申请人： Kevin J. Ash ,  Lokesh M. Gupta ,  Thomas C. Jarvis ,  Steven R. Lowe

发明人： Kevin J. Ash ,  Lokesh M. Gupta ,  Thomas C. Jarvis ,  Steven R. Lowe

IPC分类号： G06F13/00

CPC分类号： G06F12/0804 ,  G06F11/1076 ,  G06F11/108 ,  G06F12/0866 ,  G06F2211/1009 ,  G06F2211/1057 ,  G06F2211/1059

摘要： A method, system and processor for substantially reducing the write penalty (or latency) associated with writes and/or destaging operations within a RAID 5 array and/or RAID 6 array. When a write or destaging operation is initiated, i.e., when modified data is to be evicted from the cache, an existing data selection mechanism first selects the track of data to be evicted from the cache. The data selection mechanism then triggers a data track grouping (DTG) utility, which executes a thread to group data tracks, in order to maximize full stripe writes. Once the DTG algorithm completes the grouping of data tracks to complete a full stripe, a fall-stripe write is performed, and parity is generated without requiring a read from the disk(s). In this manner, the write penalty is substantially reduced, and the overall write performance of the processor is significantly improved.

摘要翻译： 一种用于显着减少与RAID 5阵列和/或RAID 6阵列内的写入和/或降级操作相关联的写入损失（或等待时间）的方法，系统和处理器。 当启动写入或降级操作时，即当修改的数据要从缓存中逐出时，现有的数据选择机制首先从缓存中选择要被驱逐的数据的轨道。 然后，数据选择机制触发数据轨道分组（DTG）实用程序，该实用程序执行线程来对数据轨道进行分组，以便最大限度地实现全部条带写入。 一旦DTG算法完成了数据轨道的分组以完成一个完整的条带，就执行一个倒带写入，并且产生奇偶校验，而不需要从一个或多个磁盘读取。 以这种方式，大大降低了写入损失，并且显着提高了处理器的整体写入性能。

公开(公告)号：US07191207B2

公开(公告)日：2007-03-13

申请号：US10459645

申请日：2003-06-11

申请人： Lawrence C. Blount ,  Thomas C. Jarvis ,  Steven R. Lowe ,  David F. Mannenbach

发明人： Lawrence C. Blount ,  Thomas C. Jarvis ,  Steven R. Lowe ,  David F. Mannenbach

IPC分类号： G09F15/173 ,  G06F9/00

CPC分类号： G06F3/0613 ,  G06F3/0659 ,  G06F3/0689 ,  G06F12/0866

摘要： An apparatus and a method to allocate device adapter bandwidth, where that device adapter is disposed in an information storage and retrieval system capable of communicating with one or more host computers, and where that device adapter interconnects a data cache with information storage devices. The data storage and retrieval system receives a plurality of requests from the host computers to read information from, or to write information to, information storage devices. The method determines a read/write request ratio and a non-volatile storage utilization ratio. The method allocates device adapter bandwidth based upon the read/write request ratio and/or the non-volatile storage utilization ratio.

摘要翻译： 一种分配设备适配器带宽的设备和方法，其中该设备适配器被布置在能够与一个或多个主机计算机通信的信息存储和检索系统中，并且该设备适配器将数据高速缓存与信息存储设备互连。 数据存储和检索系统从主计算机接收多个请求以从信息存储装置读取信息或将信息写入信息存储装置。 该方法确定读/写请求比率和非易失性存储利用率。 该方法基于读/写请求比和/或非易失性存储利用率分配设备适配器带宽。