Optimizing a file system for different types of applications in a compute cluster using dynamic block size granularity
    1.
    发明授权
    Optimizing a file system for different types of applications in a compute cluster using dynamic block size granularity 有权
    使用动态块大小粒度优化计算集群中不同类型应用程序的文件系统

    公开(公告)号:US09021229B2

    公开(公告)日:2015-04-28

    申请号:US12760476

    申请日:2010-04-14

    IPC分类号: G06F3/06

    摘要: Embodiments of the invention relate to optimizing a file system for different types of applications in a compute cluster using dynamic block size granularity. An exemplary embodiment includes reserving a predetermined number of storage allocation regions for each node in a cluster, wherein each storage allocation region comprises a set of contiguous data blocks on a single storage disk of locally attached storage; using a contiguous set of data blocks on a single storage disk of locally attached storage as a file system's block allocation granularity for striping data to be stored in the file system for a compute operation in the cluster; and using a single data block of a shared storage subsystem as the file system's block allocation granularity for striping data to be stored in the file system for a data access operation in the cluster.

    摘要翻译: 本发明的实施例涉及使用动态块大小粒度来优化在计算集群中的不同类型的应用的文件系统。 示例性实施例包括为集群中的每个节点保留预定数量的存储分配区域,其中每个存储分配区域包括在本地连接的存储器的单个存储盘上的一组连续的数据块; 在本地连接的存储的单个存储盘上使用连续的一组数据块作为文件系统的块分配粒度,用于条带化要存储在文件系统中的数据以用于群集中的计算操作; 并且使用共享存储子系统的单个数据块作为文件系统的块分配粒度,用于条带化要存储在文件系统中的数据以用于群集中的数据访问操作。

    Optimizing a File System for Different Types of Applications in a Compute Cluster Using Dynamic Block Size Granularity
    2.
    发明申请
    Optimizing a File System for Different Types of Applications in a Compute Cluster Using Dynamic Block Size Granularity 有权
    使用动态块大小粒度优化计算群集中不同类型应用程序的文件系统

    公开(公告)号:US20110258378A1

    公开(公告)日:2011-10-20

    申请号:US12760476

    申请日:2010-04-14

    IPC分类号: G06F12/00 G06F12/02

    摘要: Embodiments of the invention relate to optimizing a file system for different types of applications in a compute cluster using dynamic block size granularity. An exemplary embodiment includes reserving a predetermined number of storage allocation regions for each node in a cluster, wherein each storage allocation region comprises a set of contiguous data blocks on a single storage disk of locally attached storage; using a contiguous set of data blocks on a single storage disk of locally attached storage as a file system's block allocation granularity for striping data to be stored in the file system for a compute operation in the cluster; and using a single data block of a shared storage subsystem as the file system's block allocation granularity for striping data to be stored in the file system for a data access operation in the cluster.

    摘要翻译: 本发明的实施例涉及使用动态块大小粒度来优化在计算集群中的不同类型的应用的文件系统。 示例性实施例包括为集群中的每个节点保留预定数量的存储分配区域,其中每个存储分配区域包括在本地连接的存储器的单个存储盘上的一组连续的数据块; 在本地连接的存储的单个存储盘上使用连续的一组数据块作为文件系统的块分配粒度,用于条带化要存储在文件系统中的数据以用于群集中的计算操作; 并且使用共享存储子系统的单个数据块作为文件系统的块分配粒度,用于条带化要存储在文件系统中的数据以用于群集中的数据访问操作。

    Fileset masks to cluster inodes for efficient fileset management
    3.
    发明授权
    Fileset masks to cluster inodes for efficient fileset management 有权
    文件集掩码到集群inode以进行高效的文件集管理

    公开(公告)号:US09002911B2

    公开(公告)日:2015-04-07

    申请号:US12847937

    申请日:2010-07-30

    IPC分类号: G06F7/00 G06F17/30

    CPC分类号: G06F17/30091 G06F17/30224

    摘要: A computer-implemented method enables efficient tracking of fileset IDs and inode utilizing a fileset mask and an internal inode mask. A utility dynamically updates the fileset mask and the internal inode mask to enable mapping of the number of one (1) bits in the fileset mask to a fileset identifier (ID) and mapping of the one bits in the inode mask to an inode number for the fileset, as either the number of filesets or the number of inodes required per filesets increases above a next power of two threshold. The one bits within the fileset mask and the internal inode mask are disjoint relative to each other. An external inode number is mapped to a corresponding fileset and to a corresponding inode number by concatenating individual bits of the external inode number corresponding to each one (1) bit within the fileset mask and the inode mask.

    摘要翻译: 计算机实现的方法使得能够利用文件集掩码和内部inode掩码来有效地跟踪文件集ID和节点。 实用程序动态更新文件集掩码和内部inode掩码,以便将文件集掩码中的一(1)位的数量映射到文件集标识符(ID),并将索引节点掩码中的一位映射到 文件集作为文件集的数量或每个文件集所需的inode数量增加到下一个两个阈值的幂。 文件集掩码和内部inode掩码中的一个位相对于彼此是不相交的。 通过连接与文件集掩码中的每一(1)位对应的外部索引节点号的独立位和索引节点掩码,将外部inode号映射到相应的文件集和相应的inode号。