公开(公告)号：JP2007184011A

公开(公告)日：2007-07-19

申请号：JP2007104593

申请日：2007-04-12

Applicant: Network Appliance Inc ,  ネットワーク・アプライアンス・インコーポレイテッド

Inventor： HITZ DAVID ,  MALCOLM MICHAEL ,  LAU JAMES ,  RAKITZIS BYRON

IPC: G06F3/06 ,  G06F12/16 ,  G06F11/10 ,  G11B20/18

CPC classification number: G06F11/1076 ,  G06F2211/1059 ,  G06F2211/1061 ,  G11B20/1833

Abstract: PROBLEM TO BE SOLVED: To provide a system for shortening time necessary for recalculating a parity after the system fails and for preventing data in a RAID array from being destructed. SOLUTION: The storage subsystem is provided with: a RAID array including a plurality of disks for recording a plurality of stripe data and parity information; a non-volatile memory independent from the disks; and a RAID controller for responding to a write processing request by storing stripe information including information for specifying a stripe in the RAID array in the non-volatile memory. The stripe information includes stripe number information, and the RAID controller recalculates only a parity for the specific stripe by using the stripe information after system failure generated before the completion of write processing and restores the RAID array to a suitable state without recalculating the parities of other stripes. COPYRIGHT: (C)2007,JPO&INPIT

Abstract translation: 要解决的问题：提供一种系统，用于缩短在系统故障之后重新计算奇偶校验所需的时间并且防止RAID阵列中的数据被破坏。

解决方案：存储子系统具有：RAID阵列，其包括用于记录多个条带数据和奇偶校验信息的多个盘; 独立于磁盘的非易失性存储器; 以及RAID控制器，用于通过将包括用于在RAID阵列中指定条带的信息的条带信息存储在非易失性存储器中来响应写处理请求。 条带信息包括条带号信息，并且RAID控制器仅在写入处理完成之前通过使用生成系统故障之后的条带信息重新计算特定条带的奇偶校验，并将RAID阵列恢复到合适的状态，而不重新计算其他条目的奇偶校验 条纹。 版权所有（C）2007，JPO＆INPIT

公开(公告)号：JP2008251034A

公开(公告)日：2008-10-16

申请号：JP2008133346

申请日：2008-05-21

Applicant: Network Appliance Inc ,  ネットワーク アプライアンス, インコーポレイテッド

Inventor： HITZ DAVID ,  MALCOLM MICHAEL ,  LAU JAMES ,  RAKITZIS BYRON

IPC: G06F3/06 ,  G06F12/16 ,  G06F11/10 ,  G11B20/18

CPC classification number: G06F11/1076 ,  G06F2211/1059 ,  G06F2211/1061 ,  G11B20/1833

Abstract: PROBLEM TO BE SOLVED: To provide a system for shortening time necessary for recalculating a parity after the system fails and for preventing data in a RAID array from being destructed. SOLUTION: The storage subsystem is provided with: a RAID array including a plurality of disks for recording a plurality of stripe data and their parity values; a non-volatile memory for storing immediately preceding stripe data to be updated by a write processing; and a RAID controller configured to store an existing immediately preceding stripe data stored in the non-volatile memory before the write processing is performed. The RAID controller calculates an update parity information to the destructed specified stripe by using the existing data in the non-volatile memory to the data block of the stripe determined to be indecipherable due to the failure of the system and the data read out from the data block of the decipherable stripe. COPYRIGHT: (C)2009,JPO&INPIT

Abstract translation: 要解决的问题：提供一种系统，用于缩短在系统故障之后重新计算奇偶校验所需的时间并且防止RAID阵列中的数据被破坏。

解决方案：存储子系统具有：RAID阵列，包括用于记录多个条带数据及其奇偶校验值的多个盘; 用于存储要通过写入处理更新的紧接在前的条带数据的非易失性存储器; 以及RAID控制器，被配置为在执行写入处理之前存储存储在非易失性存储器中的现有的紧接在前的条带数据。 RAID控制器通过使用非易失性存储器中的现有数据，由于系统故障和从数据读出的数据被确定为不可解密的条带的数据块来计算更新奇偶校验信息 块的可解密条带。 版权所有（C）2009，JPO＆INPIT

公开(公告)号：JP2006260582A

公开(公告)日：2006-09-28

申请号：JP2006107735

申请日：2006-04-10

Applicant: Network Appliance Inc ,  ネットワーク・アプライアンス・インコーポレイテッド

Inventor： HITZ DAVID ,  MALCOLM MICHAEL ,  LAU JAMES ,  RAKITZIS BYRON

IPC: G06F3/06 ,  G06F11/10 ,  G06F12/00 ,  G06F17/30

CPC classification number: G06F11/1076 ,  G06F3/0601 ,  G06F3/0613 ,  G06F3/0643 ,  G06F3/0689 ,  G06F2003/0697 ,  Y10S707/99955

Abstract: PROBLEM TO BE SOLVED: To integrate a file system with a RAID array so as to export precise information about the arrangement of data blocks in the RAID subsystem. SOLUTION: Individual CWL pointers are used for each disk 1022 in the disk array 1030. The algorithm aims to improve efficiency of RAID and read-out performance. For the above purpose, it always writes on the disk 1022 with the lowest CWL pointer. In addition, another disk is chosen only when space for a new file is start to be allocated or when N blocks on the same disk 1022 has been allocated for a single file. The result is that CWL pointers are never more than N blocks apart even on different disks 1024, and a large file can also be allocated in N consecutive blocks on the same disk 1022. COPYRIGHT: (C)2006,JPO&NCIPI

Abstract translation: 要解决的问题：将文件系统与RAID阵列集成，以便导出关于RAID子系统中数据块排列的精确信息。 解决方案：单个CWL指针用于磁盘阵列1030中的每个磁盘1022.该算法旨在提高RAID和读出性能的效率。 为了上述目的，它始终以最低的CWL指针写在盘1022上。 此外，仅当新文件的空间开始分配时，或者当同一磁盘1022上的N个块已被分配给单个文件时，才选择另一个磁盘。 结果是即使在不同的磁盘1024上，CWL指针也不会超过N个块，并且大的文件也可以在同一个磁盘102上的N个连续块中分配。（C）2006年，JPO和NCIPI

公开(公告)号：WO9945456B1

公开(公告)日：1999-11-11

申请号：PCT/US9904550

申请日：1999-03-02

Applicant: NETWORK APPLIANCE INC

Inventor： HITZ DAVID ,  BORR ANDREA ,  HAWLEY ROBERT J ,  MUHLESTEIN MARK ,  PEARSON JOAN

IPC: G06F12/00 ,  G06F1/00 ,  G06F21/41 ,  G06F21/62

CPC classification number: G06F21/6218 ,  G06F21/41 ,  G06F21/6236 ,  G06F2221/2141

Abstract: The invention provides a method and system for enforcing file access control among client devices using multiple diverse access control models and multiple diverse file server protocols. A multi-protocol file server identifies each file with one particular access control model out of a plurality of possible models, and enforces that one particular model for all accesses to that file. When the file server receives a file server request for that file using a different access control model, the file server translates the access control limits for that file into no-less-restrictive limits in the different model. The file server restricts access by the client device using the translated access control limits. Each file is assigned the access control model of the user who created the file or who last set access control limits for the file. When a user having a different access control model sets access control limits, the access control model for the file is changed to the new model. Files are organized in a tree hierarchy, in which each tree is limited to one or more access control models (which can limit the ability of users to set access control limits for files in that tree). Each tree can be limited to NT-model-only format, Unix-model-only format, or mixed NT-or-Unix-models format.

Abstract translation: 本发明提供一种用于使用多种不同的访问控制模型和多种多样的文件服务器协议来实施客户端设备之间的文件访问控制的方法和系统 多协议文件服务器使用多个可能模型中的一个特定的访问控制模型识别每个文件，并对该文件的所有访问执行该特定模型。 当文件服务器使用不同的访问控制模型接收到该文件的文件服务器请求时，文件服务器将该文件的访问控制限制转换为不同模型中的无限制限制。 文件服务器使用翻译的访问控制限制来限制客户端设备的访问。 每个文件都分配了创建文件的用户的访问控制模型，或者最后为文件设置访问控制限制的访问控制模型。 当具有不同访问控制模型的用户设置访问控制限制时，将文件的访问控制模型更改为新模型。 文件以树状层次结构组织，其中每个树被限制到一个或多个访问控制模型（这可以限制用户为该树中的文件设置访问控制限制的能力）。 每个树可以限制为仅NT型格式，仅限Unix模型格式或混合的NT或Unix模型格式。

公开(公告)号：WO2006088773A3

公开(公告)日：2006-10-12

申请号：PCT/US2006004943

申请日：2006-02-13

Applicant: NETWORK APPLIANCE INC

Inventor： HITZ DAVID ,  EDWARDS JOHN K

IPC: G06F3/06

CPC classification number: G06F3/0689 ,  G06F3/0607 ,  G06F3/0643 ,  Y10S707/99956

Abstract: A system and method enables a storage system to support multiple volume type simultaneously. A volume type field is contained within a file system information block that permits the storage system to determine the type of volume of a particular volume associated therewith. The storage operating system may then interpret various on-disk data structures in accordance with the appropriate volume type.

Abstract translation: 系统和方法使存储系统能够同时支持多种卷类型。 卷类型字段包含在文件系统信息块中，其允许存储系统确定与其相关联的特定卷的体积的类型。 然后，存储操作系统可以根据适当的卷类型来解释各种在盘数据结构。

公开(公告)号：CA2322320C

公开(公告)日：2010-02-09

申请号：CA2322320

申请日：1999-03-02

Applicant: NETWORK APPLIANCE INC

Inventor： HITZ DAVID ,  BORR ANDREA ,  HAWLEY ROBERT J ,  MUHLESTEIN MARK ,  PEARSON JOAN

IPC: G06F12/00 ,  G06F1/00 ,  G06F21/41 ,  G06F21/62

Abstract: The invention provides a method and system for enforcing file access control among client devices using multiple diverse access control models and multiple diverse file server protocols. A multi-protocol file server identifies each file with one particular access control model out of a plurality of possible models, and enforces that one particular model for all accesses to that file. When the file server receives a file server request for that file using a different access control model, the file server translates the access control limits for that file into no-less-restrictive limits in the different model. The file server restricts access by the client device using the translated access control limits. Each file is assigned the access control model of the user who created the file or who last set access control limits for the file. When a user having a different access control model sets access control limits, the access control model for the file is changed to the new model. Files are organized in a tree hierarchy, in which each tree is limited to one or more access control models (which can limit the ability of users to set access control limits for files in that tree). Each tree can be limited to NT-model-only format, Unix-model-only format, or mixed NT-or-Unix-models format.

公开(公告)号：HK1013871A1

公开(公告)日：1999-09-10

申请号：HK98115311

申请日：1998-12-24

Applicant: NETWORK APPLIANCE INC

Inventor： HITZ DAVID ,  MALCOLM MICHAEL ,  LAU JAMES ,  RAKITZIS BYRON

IPC: G06F3/06 ,  G06F11/10 ,  G06F12/00 ,  G06F17/30 ,  G06F

Abstract: The present invention is a method for integrating a file system with a RAID array (1030) that exports precise information about the arrangement of data blocks in the RAID subsystem (1030). The system uses explicit knowledge of the underlying RAID disk layout to schedule disk allocation. The present invention uses separate current-write location (CWL) pointers for each disk (1022) in the disk array (1030) where the pointers simply advance through disks (1022) as writes occur. The algorithm used has two primary goals. The first goal is to keep the CWL pointers as close together as possible, thereby improving RAID (1030) efficiency by writing to multiple blocks in the stripe simultaneously. The second goal is to allocate adjacent blocks of a file on the same disk (1022), thereby improving read back performance. The first goal is satisfied by always writing on the disk (1022) with the lowest CWL pointer. For the second goal, another disk (1024) is chosen only when the algorithm starts allocating space for a new file, or when it has allocated N blocks on the same disk (1022) for a single file. The result is that CWL pointers are never more than N blocks apart on different disks (1024), and large files have N consecutive blocks on the same disk (1022).

公开(公告)号：DE60333648D1

公开(公告)日：2010-09-16

申请号：DE60333648

申请日：2003-06-03

Applicant: NETWORK APPLIANCE INC

Inventor： HITZ DAVID ,  EDWARDS JOHN ,  LEWIS BLAKE

IPC: G06F12/00 ,  G06F17/30

公开(公告)号：DE69929772D1

公开(公告)日：2006-04-20

申请号：DE69929772

申请日：1999-03-02

Applicant: NETWORK APPLIANCE INC

Inventor： HITZ DAVID ,  BORR ANDREA ,  HAWLEY J ,  MUHLESTEIN MARK ,  PEARSON JOAN

IPC: G06F1/00 ,  G06F12/00 ,  G06F21/41 ,  G06F21/62

公开(公告)号：DE60025749D1

公开(公告)日：2006-04-13

申请号：DE60025749

申请日：2000-10-25

Applicant: NETWORK APPLIANCE INC

Inventor： KLEIMAN ROBERT ,  HITZ DAVID ,  HUTCHINSON C ,  O'MALLEY W

IPC: G06F11/14 ,  G06F12/00 ,  G06F11/20 ,  G06F17/30