JP2008033921A - Distributed storage system with accelerated striping - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an intelligent data storage sub system for providing a capacity as a virtual storage space to a network in order to conduct self-determining allocation of respective data storage capacities, manage, and protect the respective data storage capacities, and store a total storage request. <P>SOLUTION: A data storage device is provided with a resident software system in a memory space configured so as to encode data read from a first number of logical units into a single channel in order to store the data in a second number of logical units. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

(Related application)
This application is a continuation-in-part of US patent application Ser. No. 11 / 145,403, filed Jun. 3, 2005, assigned to the assignee of the present application.

  The present invention relates generally to the field of distributed data storage systems, and more particularly, but not exclusively, to methods and apparatus for restriping data stored in memory.

  Computer networks began to proliferate when the data transfer rate of industry standard architectures could not match the access rate of the 80386 processor manufactured by Intel. The local area network (LAN) has evolved into a storage area network (SAN) by consolidating data storage capabilities into the network. Aggregation of relevant data and equipment processed by SAN equipment, such as the ability to handle larger storage beyond the order of storage directly connected, allows users to realize significant benefits and We were able to do it at a manageable cost.

  Recently, there has been a trend toward network-centric methods for controlling data storage subsystems. That is, the system that controls the functions of the storage unit is also separated from the server and the network itself, just as the storage unit is integrated. For example, host-based software can delegate maintenance and management tasks to intelligent switches or special network storage service platforms. Appliance-based solutions do not require software to run on the host and operate in computers that are deployed as enterprise nodes. In any case, intelligent network solutions can centralize things like storage allocation routines, backup routines, and fault tolerant schemes independently of the host.

  While the migration of intelligence from the host to the network has solved some of these problems, it does not solve the inherent difficulties because it generally lacks the flexibility to change the representation of virtual memory to the host. For example, stored data needs to be moved due to reliability issues, or more storage capacity needs to be added to accommodate the growing network. In these cases, either the host or the network must be changed to signal the presence of the new or changed storage space. What is needed is an intelligent data storage subsystem that self-determinably allocates, manages and protects its respective data storage capacity and provides that capacity as virtual storage space to the network to accommodate the entire storage request . A distributed computing environment uses these intelligent storage devices along with global provisioning for global striping and restriping of stored data. It is this solution to which embodiments of the present invention are directed.

  Embodiments of the present invention are generally directed to distributed storage systems that have a re-striping capability that avoids the intensive overhead associated with backing up originally stored data and restores it to a restripe configuration.

  In one embodiment, a software system residing in a memory space configured to encode data read from a first number of logical devices into a single channel for storing data in the second number of logical devices. Provided for data storage.

  In one embodiment, a method is provided for encoding data striped on a first number of logical devices into a single channel and decoding the encoded data into a second number of logical devices.

  In one embodiment, an intelligent storage element is provided to a data storage system having a memory space and having a device that restripes data to logical devices in the memory space.

  These and various other features and advantages that characterize the claimed invention will become apparent upon reading the following detailed description and upon reference to the associated drawings.

  FIG. 1 is an exemplary computer system 100 in which embodiments of the present invention are useful. One or more hosts 102 are networked to one or more network connection servers 104 via a local area network (LAN) and / or a wide area network (WAN) 106. The LAN / WAN 106 preferably uses an Internet Protocol (IP) network infrastructure that communicates through the World Wide Web. The host 102 accesses applications residing in the server 104 that routinely require data stored in one or more of a number of intelligent storage elements (“ISE”) 108. Accordingly, SAN 110 connects server 104 to ISE 108 to access stored data. The ISE 108 provides a block of data storage capacity 109 for storing data through various selective communication protocols such as Serial ATA and Fiber Channel, having enterprise or desktop class storage media therein.

  FIG. 2 is a simplified diagram of the computer system of FIG. Hosts 102 interact with each other and with a pair of ISEs 108 (denoted A and B, respectively) via a network or fabric 110. Each ISE 108 includes a dual redundant controller 112 (denoted A1, A2 and B1, B2) that preferably operates on the data storage capacity 109 as a set of data storage devices characterized as a redundant array (RAID) of independent drives. Including. The controller 112 and the data storage capacity 109 use a redundant link in which the various controllers 112 use parallel, and at least some of the user data stored by the system 100 is stored in at least one set of the data storage capacity 109. It is desirable to use a fault tolerant arrangement.

  A host computer 102 and A ISE 108 are physically located at the first site, B host computer 102 and B ISE 108 are physically located at the second site, and C host computer 107 is further at the third site. It is further believed that this is possible, but this is merely exemplary and not limiting. All entities on a distributed computer system are connected through some type of computer network.

  FIG. 3 illustrates an ISE 108 configured in accordance with an embodiment of the present invention. The shelf 114 provides a cavity that receptively engages the controller 112 that is electrically connected to the midplane 116. The shelf 114 is also supported in a cabinet (not shown). A multiple disk assembly (MDA) 118 pair is receptively engaged with the shelf 114 on the same side of the midplane 116. On the opposite side of the midplane 116, a dual battery 122, a dual AC power supply 124, and a dual interface module 126 that provide emergency power are connected. It is desirable that the dual components be configured to operate on either the MDA or both at the same time to provide backup protection in the event of component failure.

  FIG. 4 is an enlarged partial isometric view of MDA 118 constructed in accordance with an embodiment of the present invention. The MDA 118 has an upper section 130 and a lower section 132, each supporting five data storage devices 128. The compartments 130, 132 are aligned with a data storage device 128 that connects to a common circuit board 134 having a connector 136 that operatively engages the midplane 116 (FIG. 3). The wrapper 138 provides an electromagnetic interference shield. This exemplary embodiment of MDA 118 is the subject matter of Japanese Patent Application No. 10 / 884,605, which is assigned to the assignee of the present invention and is incorporated herein by reference. is there. Another exemplary embodiment of MDA is the subject matter of Japanese Patent Application No. 10 / 817,378 of the same title assigned to the assignee of the present invention and incorporated herein by reference. In another equivalent embodiment, the MDA 118 can be provided in a sealed housing as described below.

  FIG. 5 is an isometric view of an exemplary data storage device 128 suitable for use with embodiments of the present invention, in the form of a rotating media disk drive. Although a rotating spindle with a moving data storage medium is used in the following description, another equivalent embodiment uses a non-rotating media device such as a solid-state memory element. The data storage disk 140 is rotated by a motor 142 to provide the data storage position of the disk 140 to a read / write head (“head”) 143. The head 143 is supported by the distal end of a rotary actuator 144 that can move the head 143 radially between the internal and external tracks of the disk 140. Head 143 is electrically connected to circuit board 145 via flex circuit 146. Circuit board 145 is adapted to transmit and receive control signals that control the function of data storage device 128. A connector 148 is electrically connected to the circuit board 145 to connect the data storage device 128 to the circuit board 134 of the MDA 118 (FIG. 4).

  FIG. 6 is a diagram of an ISE 108 constructed in accordance with an embodiment of the present invention. Controller 112 operates in conjunction with intelligent storage processor (ISP) 150 to provide managed reliability of data integrity. ISP 150 can reside anywhere within controller 112, MDA 118, or ISE 108.

  Controller 112 responds to remote access commands to data packs 151, 153 via communication ports 155, 157, respectively. For convenience of explanation, the controller 112 of FIG. 6 creates logical units (“LUN”) 1 and LUN 2 from the data pack 151, all responding to a storage capacity host command. For the convenience of this description, it is assumed that the data packs 151 and 153 include eight data storage devices 128 for data storage and two spare data storage devices 128.

  Management reliability features include invoking a reliable data storage format such as a RAID strategy. For example, by providing a system that selectively uses a selected one of a plurality of different RAID formats, a relatively more robust system for data storage is generated and used to manage the MDA 118. Enables optimization of firmware algorithms that reduce software complexity and enables relatively quick recovery from memory failure conditions. These and other features of this multiple RAID format system are assigned to this assignee and described in Japanese Patent Application No. 10 / 817,264, entitled Storage Medium Data Structure and Method, which is incorporated herein by reference. Has been.

  Administrative reliability also includes the scheduling of diagnostic and correction routines based on system monitoring usage. Data restoration operations are performed for data copying and reconstruction. ISP 150 is integrated into MDA 118 to facilitate “self-healing” of the total data storage capacity without data loss. These and other features of the management reliability features considered herein are assigned to this assignee and are hereby incorporated by reference into Japanese Patent Application No. 10 entitled Management Reliability Storage System and Method. / 817,617. Another feature of management reliability is, for example, Japanese Patent Application No. 11 / 040,410 entitled Definitive Preventive Recovery from Predictive Failure of a Distributed Storage System Assigned to the Assignee and incorporated herein by reference. In response to preventive failure indications associated with a given rule.

  FIG. 7 is a diagram of an ISP circuit board 152 in which a pair of redundant ISPs 150 resides. ISP 150 interfaces data storage capacity 109 to SAN fabric 110. Each ISP 150 can manage various storage services such as routing, volume management, and data migration and replication. ISP 150 divides plate 152 into two ISP subsystems 154 and 156 that are coupled by bus 158. ISP subsystem 154 includes ISP 150 labeled “B” coupled to fabric 110 and storage capacity 109 by links 160, 162, respectively. ISP 154 also includes a policy processor 164 that executes a real-time operating system. ISP 150 and policy processor 164 communicate via bus 166 and both communicate with memory 168.

  FIG. 8 is a diagram of an exemplary ISP subsystem 154 configured in accordance with an embodiment of the present invention. ISP 150 includes a number of function controllers (170-180) that communicate with list managers 182, 184 via an intersection switch (CSP) 186 message crossbar. Accordingly, the controller (170-180) generates a CSP message in response to each specific condition to access the memory module and / or initiate an ISP operation, and this message is sent through the CPS 186 to the list manager 182. , 184 can be transmitted. Similarly, responses from list managers 182, 184 can communicate with any of the controllers (170-180) via CPS 186. The description associated with the arrangement of FIG. 8 is exemplary and is not intended to limit the possible embodiments of the present invention.

  Policy processor 164 is programmable through ISP 150 to perform the required operations. For example, policy processor 164 can communicate with list managers 182, 184 that send and receive messages via CPS 186. The response to the policy processor 164 can serve as an interrupt to a signal that reads the memory 168 register.

  FIG. 9 illustrates the flexibility advantage of ISE 108 with intelligent controller 112 for communicating with host 102 over some of a plurality of preselected communication protocols, such as FC, iSCSI, or SAS. FIG. The ISE 109 is programmable to determine the level of abstraction of the host command and thereby map the virtual storage volume to the physical storage 109 associated with the command.

  For present purposes, the term “virtual storage volume” means a logical entity that generally corresponds to a logical abstraction of physical storage. A “virtual storage volume” can include entities that are processed (logically) as if they were, for example, a fixed block architecture with a count key data architecture or a block that is sequentially addressed with a record It is. A virtual storage volume can be physically located on one or more storage elements.

FIG. 10 is a diagram of a type of data management service that can be executed by an ISE 108 independent of any host 102. For example, RAID management can be controlled locally for fault tolerant data integrity by striping data performed within the required number of data storage devices 128 ₁ , 128 ₂ , 128 ₃ ,... 128 _n . . It is possible to locally control the virtualization service to allocate and / or reverse allocate memory capacity to logical entities. Application routines such as the management reliability method described above and data migration between logical volumes within the same ISE 108 can be controlled locally as well.

  In FIG. 11, the ISE 108 has created another LUN 3 in response to a host command with additional storage capacity. In this embodiment, an acceleration method is considered in which data already existing in LUN1 and LUN2 is re-striped on all three LUNs. Advantageously, the accelerated restriping of this embodiment does not require a relatively time consuming process with a related art solution that backs up existing data and then restores it to the newly configured storage space. That is, in this solution, the ISE 108 is available to the system 100 based on only when the re-stripe process is initiated and the rate of restriping, according to the process described herein. The time started and the rate at which data is transferred can be varied in relation to other system 100 resource requirements in order to minimize the negative impact of the restriping process on system 100 performance. Also, by simplifying input to the restriping application and offloading the application from the remote host 102, the present embodiment as a whole accelerates the process with less overhead spent on the distributed system 100.

  FIG. 12 shows that both LUN1 and LUN2 are filled with storage data for convenience of explanation. Columns AH represent each of the data storage devices 128 (or “domains”) in the data pack 151. Each column and row intersection, such as named “A1”, represents a chunk of storage capacity allocated to each of the domains. Multiple chunks in each row form a stripe of storage capacity on all domains. Therefore, the first stripe of LUN1 is composed of chunks A1-H1, and the first stripe of LUN2 is composed of chunks A5-H5. The space between the end of LUN1 and the start of LUN2 is for illustration only. It is assumed that there is virtually no gap in the storage space between these or other consecutive LUNs.

  FIG. 13 illustrates a software (or firmware) system 199 configured to read data from a plurality of selected LUNs residing in memory and selected via path 200 and to encode the data into a single channel via path 202. Fig. 1 schematically illustrates an ISP 150 that performs Software system 199 then decodes the encoded data to restrip the data to a second plurality of logical units via path 204.

  Software system 199 preferably performs a multiplex operation (“multiplex”) 206 to encode the data and a demultiplex operation (“demultiplex”) 208 to decode the encoded data. For the convenience of the current description, the data originally stored in LUN1 and LUN2 (FIG. 12) is restriped on LUN1-3 in accordance with an embodiment of the present invention. In other words, the currently stored data is restriped from two LUNs to three LUNs. Therefore, in FIG. 13, n = 2 and m = 3.

  FIG. 14 illustrates a multiplexing-demultiplexing software control circuit that begins by reading data stripe A1-H1 from LUN1 (from cache 168) and storing it in LUN1. FIG. 15 illustrates how the multiplexing-demultiplexing circuit is synchronously continuous on the input source and output destination, and then reads the data stripe A5-H5 from the cache 168 and stores it in LUN2. To do. FIG. 16 fully illustrates how the multiplexing-demultiplexing circuit loops serially between the input source and the output destination. That is, after going through all input sources in turn, the multiplexing operation returns to LUN1 even though the demultiplexing operation has not yet gone through all output destinations.

  In FIG. 17, data stripes A6-H6 are read from cache 168 and stored in LUN1. In FIG. 18, the multiplexing-demultiplexing circuit then reads the data stripe A3-H3 from the cache 168 and turns to store it in LUN2. In FIG. 19, data stripes A7-H7 are stored in LUN3. In FIG. 20, data stripes A4-H4 are stored in LUN1. Finally, in FIG. 21, data stripes A8-H8 are stored in LUN2.

  The restriping utility illustrated in FIGS. 14-21 is basically the same process, but can be reversed by executing processes where n = 3 and m = 2. Considering the data originally stored as being in the state as shown in FIG. 21, FIG. 22 starts the restoration process by storing data stripes A1-H1 in LUN1. In FIG. 23, the multiplexing-demultiplexing circuit cycles continuously to store data stripes A5-H5 in LUN2. FIG. 24 illustrates a circuit that next stores data stripes A2-H2 in LUN1, FIG. 25 shows data stripes A6-H6 stored in LUN2, and FIG. 26 shows data stripes A3-H3 stored in LUN1. 27, data stripes A7-H7 are stored in LUN2, and in FIG. 28, data stripes A4-H4 are stored in LUN1. Finally, in FIG. 29, data stripes A8-H8 are stored in LUN2.

  FIG. 30 is a flow diagram of a method 250 for re-striping according to an embodiment of the present invention. The method 250 begins at block 252 by selecting the number n of input nodes and the number m of output nodes for the multiplex-demultiplex operation. For example, n = 2 and m = 3 for stripes of data from two LUNs to three LUNs in FIG. At block 254, data is synchronously encoded from the input node and decoded to the output node. At block 256, it is determined whether the last data stripe has been re-striped to the newly configured logical volume. If the determination at block 256 is “YES”, the method 250 ends, otherwise control passes to block 258. At block 258, the multiplexing and demultiplexing operations are continuously looped to each next node, and control returns to block 254 where the data is encoded and decoded at the next node.

  Finally, FIG. 31 is similar to FIG. 4 but includes a plurality of data storage devices 128 and a circuit board 134 contained within a sealed enclosure made from a base 190 and a cover 192 sealed and attached thereto. Sealing and engaging the data storage device 128 that forms the MDA 118A provides the user with a number of advantages including ensuring that the placement of the data storage device 128 is not altered from a predetermined optimal configuration. Such an arrangement also allows the MDA 118A manufacturer to tune the system for optimal performance and can be clearly defined to give the number, size and type of data storage 128.

  Sealed MDA 118A also allows the manufacturer to maximize the reliability and fault tolerance of the group of storage media inside. This is done by optimizing the drive with a multiple spindle arrangement. Design optimization reduces costs, increases performance, increases reliability, and generally extends the life of data in MDA 118A. In addition, the MDA 118A design itself provides almost zero rotational vibration and a high cooling efficiency environment, which is co-pending US Patent Application No. 11/145, entitled Enhanced RVI Storage Array, assigned to the assignee of the present application. This is the subject matter of No. 404. This allows internal storage media to be manufactured on a more costly basis without compromising the reliability, performance, or capacity of the MDA 118. The sealed MDA 118A is therefore not a single point of failure, but provides almost complete rotational vibration avoidance and cooling efficiency. This allows the MDA 118A design to be optimal disk media characteristics and at the same time reduce costs while enhancing reliability and performance.

  In summary, a self-contained ISE for a distributed storage system is provided that includes a plurality of rotatable spindles each supporting a storage medium adjacent to each independently operable actuator that is in data storage and read relation to the storage medium. Is done. The ISE further includes an ISP adapted to map the virtual storage volume to a plurality of media used by the remote devices of the distributed storage system.

  In one embodiment, the ISE has a plurality of spindles and a medium contained in a common sealed housing. It is desirable for an ISP to allocate memory to a virtual storage volume that stores data in a fault tolerant manner, such as a RAID methodology. Further, the ISP can implement a management reliability methodology for the data storage process that initiates an in-situ definitive preventive restoration phase in response to the observed predicted storage failure. The ISE is preferably composed of a plurality of data storage devices each having a disk stack composed of two or more disk data storage media.

  In another embodiment, the ISE includes a plurality of self-contained discrete data storage devices and an ISP that communicates with the data storage devices to abstract instructions received from a remote device, thereby associating associated memory with the ISP. A distributed storage system including The ISP is preferably adapted to map the virtual storage volume to a plurality of data storage devices used by one or more remote devices of the distributed storage system. As before, a plurality of data storage devices and media can be contained in a common sealed housing. It is desirable for an ISP to allocate memory to a virtual storage volume that stores data in a fault-tolerant manner like a RAID methodology. The ISP can further initiate an in-situ definitive preventive restoration phase in the data storage device in response to the observed predicted storage failure.

  In another embodiment, a distributed storage system is provided that includes a host and a back-end storage subsystem that communicates with the host through a network, and includes a device that virtualizes internal storage capacity independent of the host.

  The virtualization device can feature a plurality of discrete, individually accessible data storage devices. The virtualization device may be characterized by mapping a virtual block of storage capacity associated with a plurality of data storage devices. The virtualization device may be characterized by hermetically containerizing a plurality of data storage devices and associated control units. The virtualization device can be characterized by storing data in a fault-tolerant manner, such as without restrictions on RAID methodology. The virtualization device may be characterized by initiating a definitive preventive restoration phase in situ in response to the observed predictive memory failure. The virtualization device can feature a plurality of spindle data storage arrays.

  For purposes of this specification, the term “virtualization device” does not explicitly consider previously tried solutions included in system intelligence that maps to data storage space other than within each data storage subsystem. For example, a “virtualization device” does not consider the use of a storage manager that controls the functions of the data storage subsystem, nor does it consider the placement of managers or switches within a SAN fabric or within a host.

  The present embodiment is a software system resident in a memory space and configured to encode data read from a first number of logical units into a single channel for storing data in the second number of logical units. Consider further a data storage device having: The software system is preferably configured to decode the encoded data prior to storing the data in the second number of logical devices.

  For example, the software system can be configured to multiplex the originally striped data on a first number of logical units to encode the data. In this case, the software system can be configured to demultiplex the encoded data to restrip the data on the second number of logical units. The software system preferably loops data multiplexing and data demultiplexing synchronously on one or more input sources and one or more output destinations, respectively.

  The first number of logical units and the second number of logical units can also be different, i.e., in some cases the second number is greater than the first number and in other cases the first number is greater than the second number. large.

  The software system can also restore the original striping of data on the first number of logical units by multiplexing the data that has been re-striped from the second number of logical units.

  In one embodiment, the software system can reside on intelligent storage elements of a distributed storage system.

  In another embodiment, a method is provided for encoding data on a first number of logical devices and decoding the encoded data on a second number of logical devices. The encoding stage is characterized by multiplexing the data originally striped onto the first number of logical units, and the decoding stage is demultiplexing the encoded data to re-strip the data onto the second number of logical units. It is characterized by doing. The encoding and decoding steps are characterized by synchronously looping data multiplexing and data demultiplexing to each of the one or more input sources and the one or more outputs. The method may also consider restoring the original striping of data on the first number of logical devices by multiplexing data that has been re-striped from the second number of logical devices.

  In another embodiment, a data storage system is provided having an intelligent storage element having a memory space and a device for re-striping data to a logical device in the memory space. For the purposes of this description and the accompanying claims, the meaning of the phrase “re-striping device” is not limited thereto, but may be a code as described above, such as the multiplexing and demultiplexing operations described above. Obviously requires a digitization / decoding operation. The meaning of “re-striping device” clearly does not include a previously tried solution that includes backing up the original stored data and then restoring it to a new stripe arrangement.

  Although numerous features and advantages of various embodiments of the invention have been described in the foregoing description, together with details of the structure and function of the various embodiments of the invention, this detailed description is merely exemplary and particularly not restrictive. It should be understood that changes may be made to the details regarding the structure and arrangement of parts within the principles of the invention to the maximum extent indicated by the broad general meaning of the terms presented in the claims. It is. For example, specific elements may be changed according to a specific processing environment without departing from the scope and spirit of the present invention.

  Further, while the embodiments described herein are directed to data storage arrays, the claimed subject matter is not so limited, and various other arrangements may be used without departing from the scope and spirit of the claimed invention. One skilled in the art will recognize that a processing system can be used.

1 is a graphical representation of a computer system in which embodiments of the present invention are useful. FIG. 2 is a simplified schematic representation of the computer system of FIG. FIG. 3 is an exploded isometric view of an intelligent storage element configured in accordance with an embodiment of the present invention. FIG. 4 is a partially expanded isometric view of a plurality of disk arrays of the intelligent storage element of FIG. 3. 5 is an exemplary data storage device used in the multiple disk array of FIG. FIG. 4 is a functional block diagram of the intelligent storage element of FIG. 3. FIG. 4 is a functional block diagram of an intelligent storage processor circuit board of the intelligent storage element of FIG. 3. FIG. 4 is a functional block diagram of an intelligent storage processor of the intelligent storage element of FIG. 3. FIG. 4 is a functional block diagram representation of the command abstraction and associated memory mapping service performed by the intelligent storage element of FIG. 3. FIG. 4 is a functional block diagram of another example data service performed by the intelligent storage element of FIG. 3. FIG. 7 is a diagram similar to FIG. 6, but following a host command for allocation of a new storage space. FIG. 6 is a matrix representation of data stored in LUN1 and LUN2. 2 is a diagrammatic representation of a software system configured to operate in accordance with an embodiment of the present invention. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation of continuous synchronous multiplexing and demultiplexing associated with restriping of data originally stored in two LUNs on three LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. A continuous diagram representation that restores data from three LUNs to two LUNs. FIG. 5 is a flow diagram illustrating steps in a method of restriping according to an embodiment of the present invention. FIG. 5 is an exploded isometric view similar to FIG. 4 but having a data storage device and a circuit board housed in a sealed housing.

Explanation of symbols

108 ISE
109 Data storage 112 Control device 118 MDA
128 data storage device 150 ISP
154, 156 ISP Subsystem 164 Policy Processor 170-180 Function Controller 182, 184 List Manager 186 CSP
199 Software system 206 Multiplexing operation 208 Demultiplexing operation

Claims (20)

  In a data storage device, including a software system resident in a memory space configured to encode data retrieved from a first number of logical devices into a single channel and store the data in a second number of logical devices. Data storage device characterized.   The apparatus of claim 1, wherein the software system is configured to decode the encoded data before storing the data in the second number of logical devices.   3. The apparatus of claim 2, wherein the software system is configured to multiplex data originally striped on the first number of logical units to encode data.   4. The apparatus of claim 3, wherein the software system is configured to demultiplex encoded data to restrip data on the second number of logical units.   The apparatus of claim 1, wherein the first number and the second number are different.   6. The apparatus of claim 5, wherein the second number is greater than the first number.   6. The apparatus of claim 5, wherein the first number is greater than the second number.   5. The apparatus of claim 4, wherein the software system restores original striping of data on the first number of logical units by multiplexing restriped data from the second number of logical units. An apparatus characterized by being configured as follows.   5. The apparatus of claim 4, wherein the software system is configured to synchronize data multiplexing and demultiplexing continuously on each of the one or more input sources and the one or more output destinations. A device characterized by.   The apparatus of claim 9, wherein the software system is configured to serially loop between an input source and an output destination during data multiplexing and data demultiplexing, respectively.   11. The apparatus of claim 10, wherein the software system is resident on an intelligent storage element of the distributed storage system. Encoding data striped on a first number of logical units into a single channel;
Decoding the encoded data onto a second number of logic units;
Including methods.   13. The method of claim 12, wherein the encoding step multiplexes the originally striped data into a first number of logical units. 14. The method of claim 13, wherein the decoding step demultiplexes the encoded data to restrip the data over a second number of logical units.   13. The method of claim 12, wherein the first number is different from the second number.   16. The method of claim 15, wherein the first number is greater than the second number.   16. The method of claim 15, wherein the first number is less than the second number.   15. The method of claim 14, further comprising restoring original striping of data on the first number of logical devices by multiplexing restriped data from the second number of logical devices.   15. The method of claim 14, wherein the encoding step and the decoding step synchronously loop data multiplexing and data demultiplexing to one or more input sources and one or more output destinations, respectively. A method characterized by that. In a data storage device,
An intelligent storage element having a memory space;
A device that restripes data to a logical unit of memory space; and
A storage device characterized by including data.

Images