公开(公告)号：US20190340157A1

公开(公告)日：2019-11-07

申请号：US16517436

申请日：2019-07-19

Applicant: VMware, Inc.

Inventor： William EARL ,  Christos KARAMANOLIS ,  Kiran JOSHI

IPC: G06F16/17 ,  G06F11/14

Abstract: Embodiments of the disclosure provide techniques managing a log-structured solid state drive (SSD) format in a distributed storage system. SSDs in the distributed storage system maintains a journal of logical changes to storage objects to persist prepared and committed changes in the latency path. The journal includes metadata entries that describe changes and reference data pages. Dense data structures (such as a logical block addressing table) index the metadata entries. To reduce the amount of overhead in I/O operations, the distributed storage system maintains the dense data structures in memory rather than on disk.

公开(公告)号：US20160342462A1

公开(公告)日：2016-11-24

申请号：US14716756

申请日：2015-05-19

Applicant: VMware, Inc.

Inventor： Christos KARAMANOLIS ,  Wenguang WANG ,  Kiran JOSHI ,  Sandeep RANGASWAMY

IPC: G06F11/10 ,  G06F3/06 ,  H04L29/08

CPC classification number: G06F11/1004 ,  G06F3/0619 ,  G06F3/064 ,  G06F3/0664 ,  G06F3/067 ,  H04L67/1097 ,  H04L67/38

Abstract: Exemplary methods, apparatuses, and systems include a first layer of a virtual storage area network (VSAN) module receiving a write request from a data compute node. The write request includes data to be written and the VSAN module is distributed across a plurality of computers to provide an aggregate object store using storage attached to each of the plurality of computers. The first layer of the VSAN module calculates a checksum for the data to be written and passes the data to be written and the checksum to a second layer of the VSAN module. The second layer of the VSAN module calculates a first verification checksum for the data to be written. The data and the checksum are written to persistent storage in response to determining the first verification checksum matches the checksum passed by the first layer of the VSAN module.

Abstract translation: 示例性方法，装置和系统包括从数据计算节点接收写入请求的虚拟存储区域网络（VSAN）模块的第一层。 写请求包括要写入的数据，并且VSAN模块分布在多个计算机上，以使用附加到多个计算机中的每一个的存储器提供聚合对象存储。 VSAN模块的第一层计算待写入数据的校验和，并将要写入的数据和校验和传递给VSAN模块的第二层。 VSAN模块的第二层为要写入的数据计算第一个验证校验和。 响应于确定第一验证校验和与VSAN模块的第一层传递的校验和匹配，将数据和校验和写入持久存储器。

公开(公告)号：US20150370489A1

公开(公告)日：2015-12-24

申请号：US14839163

申请日：2015-08-28

Applicant: VMware, Inc.

Inventor： Faraz SHAIKH ,  Murali VILAYANNUR ,  Satyam B. VAGHANI ,  Kiran JOSHI

IPC: G06F3/06 ,  G06F9/455 ,  G06F17/30 ,  G06F12/02

CPC classification number: G06F3/0608 ,  G06F3/0631 ,  G06F3/0652 ,  G06F3/0665 ,  G06F3/067 ,  G06F9/45533 ,  G06F9/45558 ,  G06F12/0253 ,  G06F12/0261 ,  G06F12/0269 ,  G06F12/0276 ,  G06F17/30138 ,  G06F17/30194 ,  G06F17/30233 ,  G06F2009/45579 ,  G06F2212/1044 ,  G06F2212/154

Abstract: Free storage blocks previously allocated to a logical block device are released back to an underlying storage system supporting the logical block device in a manner that does not conflict with write operations that may be issued to the free storage blocks at about the same time. According to a first technique, write operations on the same storage blocks to be released are paused until the underlying storage system has completed the releasing operation or, if the write operations are issued earlier than when the underlying storage system actually performs the releasing operation, such storage blocks are not released. According to a second technique, a special file is allocated the free storage blocks, which are then made available for safe releasing.

Abstract translation: 先前分配给逻辑块设备的空闲存储块以与同时发布到空闲存储块的写入操作不冲突的方式被释放回支持逻辑块设备的底层存储系统。 根据第一技术，暂停对要释放的相同存储块的写入操作，直到底层存储系统已经完成释放操作，或者如果写入操作早于底层存储系统实际执行释放操作时发出，则 存储块未被释放。 根据第二种技术，将特殊文件分配给空闲的存储块，然后使其可用于安全释放。

公开(公告)号：US20150058291A1

公开(公告)日：2015-02-26

申请号：US14469418

申请日：2014-08-26

Applicant: VMWare, Inc.

Inventor： William EARL ,  Christos KARAMANOLIS ,  Kiran JOSHI

IPC: G06F17/30

CPC classification number: G06F16/1734 ,  G06F11/1471

Abstract: Embodiments of the disclosure provide techniques managing a log-structured solid state drive (SSD) format in a distributed storage system. SSDs in the distributed storage system maintains a journal of logical changes to storage objects to persist prepared and committed changes in the latency path. The journal includes metadata entries that describe changes and reference data pages. Dense data structures (such as a logical block addressing table) index the metadata entries. To reduce the amount of overhead in I/O operations, the distributed storage system maintains the dense data structures in memory rather than on disk.

Abstract translation: 本公开的实施例提供了在分布式存储系统中管理日志结构固态驱动（SSD）格式的技术。 分布式存储系统中的SSD维护对存储对象的逻辑更改日志，以在延迟路径中持续准备和承诺的更改。 该期刊包括描述更改和参考数据页面的元数据条目。 密集数据结构（如逻辑块寻址表）为元数据条目索引。 为了减少I / O操作的开销，分布式存储系统将密集的数据结构维护在内存而不是磁盘上。

公开(公告)号：US20130185723A1

公开(公告)日：2013-07-18

申请号：US13785794

申请日：2013-03-05

Applicant: VMWARE, INC.

Inventor： Kiran JOSHI ,  Sirishkumar RAGHURAM ,  Bich Cau LE ,  Gopala SURYANARAYANA

IPC: G06F9/455

CPC classification number: G06F9/455 ,  G06F9/45558 ,  G06F17/30194 ,  G06F2009/45579 ,  Y10S707/99933

Abstract: A method enables a central management service to operate on individual files within virtual disks associated with different virtual machines (VMs) running in one or more host computers that are each networked to a remote storage system. An IO request from a VM is received at a virtualization software layer on a host computer, wherein the IO request relates to a file stored in a virtual disk associated with the VM, and the file is individually stored in the remote storage system in accordance with a file system. The central management service is notified of the received IO request and uses the file system to access the file in the remote storage system to perform a management task on the file, and the IO request is then performed on the file by the virtualization software layer upon a notification of a successful completion of the management task.

Abstract translation: 一种方法使得中央管理服务可以在与每个联网到远程存储系统的一个或多个主机中运行的不同虚拟机（VM）相关联的虚拟磁盘内的各个文件上操作。 来自VM的IO请求在主计算机上的虚拟化软件层处被接收，其中IO请求涉及存储在与VM相关联的虚拟磁盘中的文件，并且文件被单独存储在远程存储系统中 文件系统 中央管理服务被通知接收到的IO请求，并使用文件系统访问远程存储系统中的文件，对文件执行管理任务，然后由虚拟化软件层对文件执行IO请求 通知成功完成管理任务。