公开(公告)号：US6112257A

公开(公告)日：2000-08-29

申请号：US936122

申请日：1997-09-24

申请人： Robert S. Mason, Jr. ,  Yuval Ofek ,  Natan Vishlitzky ,  Dan Arnon ,  Eitan Bachmat

发明人： Robert S. Mason, Jr. ,  Yuval Ofek ,  Natan Vishlitzky ,  Dan Arnon ,  Eitan Bachmat

IPC分类号： G06F3/06 ,  G06F3/00 ,  G06F9/00 ,  G06F11/20 ,  G06F12/00 ,  G06F13/00

CPC分类号： G06F11/2066 ,  G06F11/2069 ,  G06F2003/0697 ,  G06F3/0601

摘要： The invention relates generally to mass storage systems, and in particular to mass storage systems in which stored logical volumes are duplicated in mirrored form. The system includes a method for dynamically adjusting the mirror service policy for a disk drive system by periodically collecting statistics describing the reading and writing of data to mirrored logical volumes of the system in successive time periods and determining, from time to time, from the collected statistics, whether the mirror service policy should continue or should change. In particular, the system takes into account activity levels at the physical devices and results in more efficient accessing of logical volume pairs as well as a better balance of loading and accessing the logical volumes.

摘要翻译： 本发明一般涉及大容量存储系统，特别涉及存储逻辑卷以镜像形式复制的大容量存储系统。 该系统包括一种用于通过周期性地收集描述在连续时间段内将数据读取和写入系统的镜像逻辑卷的统计信息来统计磁盘驱动器系统的镜像服务策略的方法，并且不时地从收集的 统计，镜像服务策略是应该继续还是应该改变。 特别地，系统考虑到物理设备的活动级别，并且导致对逻辑卷对的更有效的访问以及加载和访问逻辑卷的更好的平衡。

公开(公告)号：US08990272B2

公开(公告)日：2015-03-24

申请号：US12847696

申请日：2010-07-30

申请人： Robert S. Mason, Jr. ,  David M. Shaw ,  Kevin W. Baughman ,  Stephen Fridella

发明人： Robert S. Mason, Jr. ,  David M. Shaw ,  Kevin W. Baughman ,  Stephen Fridella

IPC分类号： G06F17/30

CPC分类号： G06F17/3023

摘要： A versioned file system comprises a set of structured data representations, such as XML. Each structured data representation corresponds to a “version,” and each version comprises a tree of write-once objects rooted at a root directory manifest. Each version in the versioned file system has associated therewith a “borrow window.” When it is desired to reconstruct the file system to a point in time (or, more generally, a given state), i.e., to perform a “restore,” it is only required to walk (use) a single structured data representation (a tree). During a restore, metadata is pulled back from the cloud first, so users can see the existence of needed files immediately. The remainder of the data is then pulled back from the cloud if/when the user goes to open the file. As a result, the entire file system (or any portion thereof) can be restored to a previous time nearly instantaneously. A “fast” restore is performed if an object being restored exists within a “borrow window” of the version from which the system is restoring. A version is pruned from the versioned file system by deleting all objects in the tree (associated with the version) that, at the time of pruning: (i) are not being lent to any other version within the borrow window of the version being pruned, and (ii) are not referenced in any other version whose borrow window is sufficiently large enough such that an object in the version could have been restored from that other version.

摘要翻译： 版本化文件系统包括一组结构化数据表示，例如XML。 每个结构化数据表示对应于“版本”，并且每个版本包括一根基于根目录清单的一次写入树的树。 版本化文件系统中的每个版本都与“借用窗口”相关联。当希望将文件系统重建到一个时间点（或更一般地，给定状态）时，即执行“恢复”时， 只需要（使用）单个结构化数据表示（树）。 在还原期间，元数据首先从云中被拉回，因此用户可以立即看到所需文件的存在。 如果/当用户打开文件时，剩余的数据将从云中被拉回。 结果，整个文件系统（或其任何部分）可以几乎瞬间恢复到之前的时间。 如果正在还原的对象存在于系统还原的版本的“借用窗口”中，则执行“快速”恢复。 通过删除树中的所有对象（与版本相关联），在版本化文件系统中修剪版本：在修剪时：（i）不被借给正在修剪的版本的借用窗口内的任何其他版本 ，和（ii）在其借用窗口足够大的任何其他版本中不被引用，使得版本中的对象可能已经从该另一版本恢复。

公开(公告)号：US06895418B1

公开(公告)日：2005-05-17

申请号：US09301057

申请日：1999-04-28

申请人： Preston F. Crow ,  Robert S. Mason, Jr. ,  Steven T. McClure ,  Susan C. Nagy ,  Richard G. Wheeler

发明人： Preston F. Crow ,  Robert S. Mason, Jr. ,  Steven T. McClure ,  Susan C. Nagy ,  Richard G. Wheeler

IPC分类号： G06F17/30 ,  G06F12/00

CPC分类号： G06F17/30067 ,  Y10S707/99931 ,  Y10S707/99933 ,  Y10S707/99956

摘要： A memory storage device has a file storage operating system that uses inodes to access file segments. Each inode has a plurality of rows. A portion of the rows can store extents pointing, directly or indirectly, to data blocks. Each extent has a field to indicate whether the extent is an indirect extent or a direct extent.

摘要翻译： 存储器存储设备具有使用inode访问文件段的文件存储操作系统。 每个inode都有多行。 行的一部分可以将范围直接或间接地存储到数据块。 每个范围都有一个字段来指明范围是间接范围还是直接程度。

公开(公告)号：US06732194B1

公开(公告)日：2004-05-04

申请号：US09893295

申请日：2001-06-27

申请人： Haim Kopylovitz ,  Robert S. Mason, Jr. ,  Adi Ofer

发明人： Haim Kopylovitz ,  Robert S. Mason, Jr. ,  Adi Ofer

IPC分类号： G06F300

CPC分类号： G06F3/0659 ,  G06F3/0611 ,  G06F3/067

摘要： Described are techniques for performing multi-sequential I/O operations in connection with data requests involving a data storage device. An single data request may involve more than a one portion of data associated with a single job record, such as a single request may involve more than a single track of data of a logical device. A single job record corresponds to a single track. A data structure arrangement is disclosed that includes multiple job records corresponding to the single data request involving more than a single track of data. The multiple job records for a single data request are connected together in a data structure arrangement that may be used in connection with a single read operation involving more than a single track of data. This data structure may also be used in connection with storing a plurality of pending write requests, such as in connection with writing data from cache locations to a plurality of tracks of a particular device in which the plurality of pending write requests are represented as a single data request.

摘要翻译： 描述了与涉及数据存储设备的数据请求相关的多顺序I / O操作的技术。 单个数据请求可以涉及与单个作业记录相关联的多于一部分的数据，诸如单个请求可能涉及多于逻辑设备的单个数据轨道。 单个作业记录对应于单个轨道。 公开了一种数据结构布置，其包括与涉及多于单个数据轨迹的单个数据请求相对应的多个作业记录。 用于单个数据请求的多个作业记录以可以与涉及多于单个数据轨道的单个读取操作结合使用的数据结构布置连接在一起。 该数据结构还可以与存储多个待处理的写入请求相关联使用，例如结合将数据从高速缓存位置写入特定设备的多个轨道，其中多个待处理写入请求被表示为单个 数据请求。

公开(公告)号：US06633955B1

公开(公告)日：2003-10-14

申请号：US09965421

申请日：2001-09-27

申请人： Peng Yin ,  Robert S. Mason, Jr.

发明人： Peng Yin ,  Robert S. Mason, Jr.

IPC分类号： G06F1200

CPC分类号： G06F11/3433 ,  G06F11/2069 ,  G06F11/3485

摘要： Described are techniques for use in determining a dynamic mirror service policy (DMSP) for a plurality of mirror devices. The DMSP determines which of the plurality of mirror devices services I/O operations associated with a logical volume (LV), such as a read operation, at a particular point in time. The particular DMSP may subsequently be recalculated using device statistics from a different time interval. Part of determining a DMSP includes using device statistics to determine the activity level of each LV. The activity levels of multiple LVs may be combined to determine the activity level associated with a particular mirror device. A mirror device is selected if it has the minimum activity of all the plurality of mirror devices. Seek minimization processing is performed to minimize the distance between LVs stored on a single mirror device. Parameters used in connection with determining a DMSP may be stored in a configuration file and may be dynamically modified. Techniques described may also be used in an embodiment having a static MSP.

摘要翻译： 描述了用于确定多个镜像设备的动态镜像服务策略（DMSP）的技术。 DMSP在特定时间点确定多个镜像设备中的哪一个服务与诸如读取操作的逻辑卷（LV）相关联的I / O操作。 随后可以使用来自不同时间间隔的设备统计重新计算特定DMSP。 确定DMSP的一部分包括使用设备统计信息来确定每个LV的活动级别。 可以组合多个LV的活动水平以确定与特定镜像设备相关联的活动水平。 如果其具有所有多个镜像设备的最小活动性，则选择镜像设备。 执行最小化处理以最小化存储在单个镜像设备上的LV之间的距离。 与确定DMSP有关的参数可以存储在配置文件中，并且可以被动态修改。 所描述的技术也可以在具有静态MSP的实施例中使用

公开(公告)号：US06557079B1

公开(公告)日：2003-04-29

申请号：US09468270

申请日：1999-12-20

申请人： Robert S. Mason, Jr. ,  Yuval Ofek ,  Dan Arnon

发明人： Robert S. Mason, Jr. ,  Yuval Ofek ,  Dan Arnon

IPC分类号： G06F1200

CPC分类号： G06F12/0862

摘要： A mechanism for optimizing predictive read performance in a data storage system that is connected to a geographically remote data storage system by a data link for remote replication of data in support of data recovery operations. The data storage system initiates a local prefetch and initiates via the data link a remote prefetch by the remote data storage system to retrieve data from storage devices coupled to the local and remote data storage systems, respectively. The remote prefetch read start address is offset from the local prefetch read start address by a programmable track offset value. The programmable track offset value is adjusted to tune the prefetch workload balance between the local and remote data storage systems.

摘要翻译： 一种用于优化数据存储系统中的预测读取性能的机制，其通过数据链路连接到地理上远程的数据存储系统，用于支持数据恢复操作的数据的远程复制。 数据存储系统启动本地预取，并通过数据链路启动远程数据存储系统的远程预取，以分别从耦合到本地和远程数据存储系统的存储设备检索数据。 远程预取读取起始地址通过可编程轨道偏移值与本地预取读取起始地址偏移。 调整可编程磁道偏移值以调整本地和远程数据存储系统之间的预取工作负载平衡。

公开(公告)号：US5884098A

公开(公告)日：1999-03-16

申请号：US634419

申请日：1996-04-18

申请人： Robert S. Mason, Jr.

发明人： Robert S. Mason, Jr.

IPC分类号： G06F3/06 ,  G06F12/08 ,  G06F13/20 ,  G06F13/14

CPC分类号： G06F3/0626 ,  G06F12/0866 ,  G06F3/0635 ,  G06F3/0689

摘要： A disk drive array controller generally has a host I/O port configured for connection to a host computer and a plurality of disk I/O ports configured for connection to a corresponding plurality of disks forming a disk drive array. A controller constructed in accordance with various aspects of the present invention may include a host I/O processor in communication with the host I/O port and configured to perform I/O transactions with the host computer through the host I/O port; a cache memory; a front end caching subsystem in communication with the host I/O processor and configured to cache blocks of data comprising host I/O transactions in the cache memory; a disk array I/O processor configured to access host data in the cache memory and in communication with the plurality of disk drives, the disk array I/O processor processing host I/O transactions into disk I/O transactions; and a back end caching subsystem in communication with the disk array I/O processor, the back end caching subsystem configured to cache disk array meta-data in the cache memory. Variations of this basic system are possible, and contemplated as within the scope of the present invention. The disk drive array controller may further include a communication path between the front end caching subsystem and the back end caching subsystem, whereby allocation of blocks in the caching subsystems are synchronized. The communication path may be, for example, a control store common to the front end caching subsystem and the back end caching subsystem, the control store holding a data structure through which caching and I/O transaction information are communicated between the front end caching subsystem and the back end caching subsystem.

摘要翻译： 磁盘驱动器阵列控制器通常具有配置用于连接到主计算机的主机I / O端口和被配置用于连接到形成磁盘驱动器阵列的对应的多个磁盘的多个磁盘I / O端口。 根据本发明的各个方面构造的控制器可以包括与主机I / O端口通信并配置为通过主机I / O端口与主机执行I / O处理的主机I / O处理器; 高速缓冲存储器 前端缓存子系统，与主机I / O处理器通信并且被配置为将包括主机I / O事务的数据块缓存在高速缓冲存储器中; 磁盘阵列I / O处理器，被配置为访问高速缓冲存储器中的主机数据并与多个磁盘驱动器通信，磁盘阵列I / O处理器将主机I / O事务处理成磁盘I / O事务; 以及与所述磁盘阵列I / O处理器通信的后端缓存子系统，所述后端缓存子系统被配置为将磁盘阵列元数据缓存在所述高速缓冲存储器中。 该基本系统的变化是可能的，并且被认为在本发明的范围内。 磁盘驱动器阵列控制器还可以包括前端缓存子系统和后端高速缓存子系统之间的通信路径，由此高速缓存子系统中的块的分配被同步。 通信路径可以是例如前端缓存子系统和后端缓存子系统共用的控制存储器，控制存储器保存数据结构，通过该数据结构，在前端缓存子系统之间传送高速缓存和I / O事务信息 和后端缓存子系统。

公开(公告)号：US08661063B2

公开(公告)日：2014-02-25

申请号：US13271460

申请日：2011-10-12

申请人： Robert S. Mason, Jr. ,  David M. Shaw ,  Kevin W. Baughman ,  Stephen Fridella

发明人： Robert S. Mason, Jr. ,  David M. Shaw ,  Kevin W. Baughman ,  Stephen Fridella

IPC分类号： G06F17/30

CPC分类号： G06F17/3023 ,  G06F17/30171

摘要： A method of data sharing among multiple entities is provided. Each entity exports to a data store a structured data representation comprising a versioned file system local to that entity. The method begins by forming a sharing group that includes two or more entities. Sharing of the structured data representations by members of the sharing group is enabled. The filers use a single distributed lock to protect each version of the file system. This lock is managed to allow each filer access to the shared file system volume to create its new version. To share a fully-versioned file system, asynchronous updates at each of the filers is permitted, and each node is then allowed to “push” its individual changes to the store to form the next version of the file system. A mechanism also may be used to reduce the period during which filers in the group operate under lock.

摘要翻译： 提供了多个实体之间数据共享的方法。 每个实体向数据存储库导出包括该实体本地的版本化文件系统的结构化数据表示。 该方法从形成包含两个或多个实体的共享组开始。 启用共享组成员共享结构化数据表示。 文件管理器使用单个分布式锁来保护文件系统的每个版本。 管理此锁，以允许每个文件管理器访问共享文件系统卷以创建其新版本。 要共享完全版本的文件系统，允许每个文件管理器的异步更新，然后允许每个节点将其各个更改“推送”到存储，以形成文件系统的下一个版本。 也可以使用一种机制来减少组内的文件管理者在锁定期间的操作时间。

公开(公告)号：US08566362B2

公开(公告)日：2013-10-22

申请号：US12483030

申请日：2009-06-11

申请人： Robert S. Mason, Jr. ,  Andres Rodriguez

发明人： Robert S. Mason, Jr. ,  Andres Rodriguez

IPC分类号： G06F17/30

CPC分类号： G06F11/1446 ,  G06F11/1448 ,  G06F11/1464 ,  G06F17/30088 ,  G06F17/30132 ,  G06F17/30203 ,  G06F17/3023 ,  G06F17/30961 ,  G06F2201/805 ,  G06F2201/84

摘要： A versioned file system comprises a set of structured data representations. At a first time, an interface creates and exports to a data store a first structured data representation corresponding to a first version of a local file system. The first structured data representation is an XML tree having a root element, one or more directory elements associated with the root element, and one or more file elements associated with a given directory element. Upon a change within the file system (e.g., file creation, file deletion, file modification, directory creation, directory deletion and directory modification), the interface creates and exports a second structured data representation corresponding to a second version of the file system. The second structured data representation differs from the first structured data representation up to and including the root element of the second structured data representation. The data store may comprise a cloud storage service provider.

摘要翻译： 版本化文件系统包括一组结构化数据表示。 在第一时间，界面创建并向数据存储器导出对应于本地文件系统的第一版本的第一结构化数据表示。 第一结构化数据表示是具有根元素，与根元素相关联的一个或多个目录元素以及与给定目录元素相关联的一个或多个文件元素的XML树。 在文件系统内的变化（例如，文件创建，文件删除，文件修改，目录创建，目录删除和目录修改）之后，界面创建并输出对应于文件系统的第二版本的第二结构化数据表示。 第二结构化数据表示与第一结构化数据表示不同，并且包括第二结构化数据表示的根元素。 数据存储可以包括云存储服务提供商。

公开(公告)号：US08112423B2

公开(公告)日：2012-02-07

申请号：US13072921

申请日：2011-03-28

申请人： Benjamin K. D. Bernhard ,  Robert S. Mason, Jr.

发明人： Benjamin K. D. Bernhard ,  Robert S. Mason, Jr.

IPC分类号： G06F17/00

CPC分类号： G06F11/08 ,  G06F11/1662 ,  G06F11/2028 ,  G06F11/2094 ,  G06F11/2097

摘要： A cluster recovery process is implemented across a set of distributed archives, where each individual archive is a storage cluster of preferably symmetric nodes. Each node of a cluster typically executes an instance of an application that provides object-based storage of fixed content data and associated metadata. According to the storage method, an association or “link” between a first cluster and a second cluster is first established to facilitate replication. The first cluster is sometimes referred to as a “primary” whereas the “second” cluster is sometimes referred to as a “replica.” Once the link is made, the first cluster's fixed content data and metadata are then replicated from the first cluster to the second cluster, preferably in a continuous manner. Upon a failure of the first cluster, however, a failover operation occurs, and clients of the first cluster are redirected to the second cluster. Upon repair or replacement of the first cluster (a “restore”), the repaired or replaced first cluster resumes authority for servicing the clients of the first cluster. This restore operation preferably occurs in two stages: a “fast recovery” stage that involves preferably “bulk” transfer of the first cluster metadata, following by a “fail back” stage that involves the transfer of the fixed content data. Upon receipt of the metadata from the second cluster, the repaired or replaced first cluster resumes authority for the clients irrespective of whether the fail back stage has completed or even begun.

摘要翻译： 在一组分布式归档中实现集群恢复过程，其中每个单独的归档是优选对称节点的存储集群。 集群的每个节点通常执行提供固定内容数据和关联元数据的基于对象的存储的应用的实例。 根据存储方法，首先建立第一集群和第二集群之间的关联或“链接”以便于复制。 第一个集群有时被称为“主要”，而“第二个”集群有时被称为“副本”。一旦建立了链接，第一个集群的固定内容数据和元数据将从第一个集群复制到 第二簇，优选以连续方式。 但是，在第一个集群发生故障时，会发生故障转移操作，并将第一个集群的客户端重定向到第二个集群。 在修复或更换第一个集群（“恢复”）后，已修复或更换的第一个集群将恢复为第一个集群的客户端提供服务的权限。 该恢复操作优选地分为两个阶段：“快速恢复”阶段，其优选地涉及第一集群元数据的“批量”传送，之后是涉及传输固定内容数据的“故障恢复”阶段。 在从第二集群接收到元数据后，修复或更换的第一个集群将恢复客户端的权限，而不管故障后台是否已经完成甚至开始。