WO2014188682A1 - ストレージノード、ストレージノード管理装置、ストレージノード論理容量設定方法、プログラム、記録媒体および分散データストレージシステム - Google Patents
ストレージノード、ストレージノード管理装置、ストレージノード論理容量設定方法、プログラム、記録媒体および分散データストレージシステム Download PDFInfo
- Publication number
- WO2014188682A1 WO2014188682A1 PCT/JP2014/002562 JP2014002562W WO2014188682A1 WO 2014188682 A1 WO2014188682 A1 WO 2014188682A1 JP 2014002562 W JP2014002562 W JP 2014002562W WO 2014188682 A1 WO2014188682 A1 WO 2014188682A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- storage
- node
- storage node
- performance
- data
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/061—Improving I/O performance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0638—Organizing or formatting or addressing of data
- G06F3/0644—Management of space entities, e.g. partitions, extents, pools
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/067—Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0813—Configuration setting characterised by the conditions triggering a change of settings
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0823—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability
Definitions
- the present invention relates to a technical field for setting a storage logical capacity of a storage node.
- a distributed data system uses external storage devices of a plurality of server devices as distributed data storage nodes connected via a network.
- the distributed data storage node operates in cooperation with a plurality of node groups in which storage devices are incorporated. Therefore, the distributed data storage node is a computer system that can behave as if it were one node.
- Such a distributed data storage system may have scale-out expandability such as expansion of data storage capacity and enhancement of data access performance with the addition of nodes (that is, server devices) constituting the system. It is a feature.
- the distributed data storage system copies and stores a copy of data on a plurality of nodes instead of a single node for data requested to be written. Therefore, even if a certain node becomes inaccessible due to a failure and does not function as a storage node, the distributed data storage system can perform the following processing. In other words, in the response processing when accessing the data that was handled by the failed node, another node holding a copy of the data held by the failed node takes over the service at the failed node. . Thereby, the distributed data storage system can maintain the availability of the system.
- the distributed data storage system can use the replicated data stored in another node. For this reason, data loss does not occur, and at the same time, the reliability of data retention can be maintained.
- the distributed data storage system can be expanded in various ways in response to trends such as the ability to freely expand the system scale, the improvement of the node processing performance of the system, the increase in capacity of storage devices, and the price reduction.
- the system is being introduced.
- the distributed data storage system requires a certain number of hardware nodes depending on the service scale in the IT (Information Technology) system in order to maintain the availability of the services provided and the reliability of data retention. is necessary.
- HDDs Hard Disk Drives
- I / O performance in HDDs has been stagnant, and the I / O performance per capacity has been decreasing year by year.
- SSD Solid State Drive
- next-generation non-volatile semiconductor memories that greatly increase the capacity of volatile memories represented by DRAM (Dynamic Random Access Memory) and that significantly exceed the I / O performance of flash memories is also progressing.
- DRAM Dynamic Random Access Memory
- a plurality of storage devices having different costs, performance characteristics, and capacities have been adopted as storage device groups in a single storage system.
- the nodes constituting the distributed data storage can be combined not only with the difference in data storage capacity but also with devices with significantly different I / O performance.
- an ID identifier
- data in a certain node and a data storage destination are not managed using a data table, but data between each node and a client that accesses the data is stored.
- the client can set the storage destination independently without inquiring other nodes about the data storage destination.
- This arithmetic data storage location setting method has the following characteristics. That is, a data identifier (ID) assigned in advance to each data and the data body itself are set as input data.
- ID data identifier
- System configuration information including “storage information that associates a calculation result value with a data storage destination” is set as a parameter used for arithmetic.
- a node serving as a data storage destination is set by a value calculated using a parameter.
- System configuration information may be set so that data is uniformly distributed and stored.
- the distributed data storage system set in this way can evenly distribute the storage capacity consumption of each node and the number of data access I / Os associated with new creation and writing of data.
- Patent Document 1 proposes a multi-node storage system.
- a management node and a variety of storage nodes that manage a variety of storage devices are interconnected via a network.
- the management node acquires information related to data management from various storage nodes, and updates the logical volume as necessary.
- the system described in Patent Document 2 determines which storage node holds data based on a hash value calculated from a key assigned to the data.
- the hash value for the key can be calculated using, for example, MD5 (Message Digest algorithm 5).
- MD5 Message Digest algorithm 5
- the system may use other hash functions such as SHA (Secure Hash Algorithm).
- Japanese Patent Application Laid-Open No. 2004-228561 proposes a distributed storage system that determines an access destination node based on a node management table that manages nodes in charge of hash values.
- the system described in Patent Document 3 divides the data storage method into a first storage area and a second storage area.
- the user information is stored in the second storage area, and the user information stored in the first storage area and the user information stored in the second storage area are stored.
- Data to be replaced is distributed and arranged under conditions based on the usage frequency of user information.
- the weight for the newly installed node is set to twice that of the other nodes. As a result, it is possible to set so that data having a capacity twice as large as the probability is arranged.
- the newly introduced node needs to have I / O performance capable of processing double I / O. Otherwise, the newly introduced node becomes a performance bottleneck. Therefore, there is a problem that the performance of the entire system deteriorates.
- the storage logical capacity must be 10 times or more.
- the storage consumption rate of newly introduced nodes is higher than other nodes. As a result, the storage consumption rate of the entire system decreases.
- the distributed data storage system has a function of distributing data and data access I / O based on a method of arithmetically setting data storage destinations.
- the storage control method described in Patent Document 2 can determine an access destination node. Furthermore, the access control method described in Patent Document 3 can store data in the first storage area or the second storage area, and can exchange data according to the frequency of use of the data. .
- Patent Documents 1 to 3 cannot set the capacity of the storage area of the node based on the difference in I / O performance.
- the present invention has been made in view of such problems, and a storage node management device capable of setting a logical capacity of a storage area based on an I / O performance value of a storage node, a storage node logical capacity
- the main object is to provide a setting method, a program, a recording medium, and a distributed data storage system.
- the storage node management apparatus When a distributed data storage system is configured based on a plurality of storage nodes having different I / O performances, the storage node management apparatus according to an embodiment of the present invention is provided with at least a plurality of storage nodes having different I / O performances.
- One reference storage node is set, the I / O performance and logical capacity of the reference storage node are referred to, and storage areas of storage nodes other than the reference storage node are designated as a first storage area and a second storage area.
- And setting means for setting the logical capacities of the first and second storage areas to match the I / O performance of storage nodes other than the reference storage node; Determining means for determining system configuration information reflecting a configuration change of the distributed data storage system based on information set by the setting unit; Transmitting means for transmitting the system configuration information determined by the determining unit to the plurality of storage nodes having different I / O performances.
- the storage node logical capacity setting method when configuring a distributed data storage system based on a plurality of storage nodes having different I / O performance, the plurality of storage nodes having different I / O performances. And at least one reference storage node is set, the I / O performance and logical capacity of the reference storage node are referred to, and the storage areas of the storage nodes other than the reference storage node are designated as the first storage area and the second storage area. The storage areas are divided and the logical capacities of the first and second storage areas are set to match the I / O performance of storage nodes other than the reference storage node.
- a program according to an embodiment of the present invention may include at least one reference from the plurality of storage nodes having different I / O performances.
- the processing for setting the storage node, the I / O performance and the logical capacity of the reference storage node the storage areas of the storage nodes other than the reference storage node are changed to the first storage area and the second storage area.
- a process of dividing and setting the logical capacities of the first and second storage areas so as to match the I / O performance of the storage nodes other than the reference storage node is executed.
- a recording medium includes at least one storage node having different I / O performances. Referring to the processing for setting the reference storage node, the I / O performance and the logical capacity of the reference storage node, the storage areas of the storage nodes other than the reference storage node are designated as the first storage area and the second storage area. And a program for causing the computer to execute processing for setting the logical capacities of the first and second storage areas to match the I / O performance of storage nodes other than the reference storage node.
- a storage node includes a storage unit having a storage device that is divided into a first storage area and a second storage area by the storage management device;
- data input / output management means for determining whether the access request is for the own node; It manages a data management table that associates three types of information: data ID to be written, data storage destination physical address, and access frequency information, and the data ID and data specified by the data input / output management unit
- Data storage management means for reading and writing data in response to a read / write command
- Storage usage management means for classifying the addresses of the storage devices so as to satisfy the capacities of the first storage area and the second storage area based on the logical capacity included in the system configuration information acquired from the storage management device; Is provided.
- a distributed data storage system includes the storage management device and a storage node.
- a storage node management apparatus a storage node logical capacity setting method, a program, a recording medium, and distributed data storage that can set the logical capacity of a storage area based on the I / O performance value of the storage node A system can be provided.
- 1 is a block diagram showing a configuration of a distributed data storage system according to a first embodiment of the present invention. It is a block diagram which shows the structure of the management node which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the structure of the storage node which concerns on the 1st Embodiment of this invention.
- 3 is a flowchart showing a data write operation according to the first embodiment of the present invention.
- 4 is a flowchart showing a part of a data write operation according to the first embodiment of the present invention.
- 4 is a flowchart showing a part of a data write operation according to the first embodiment of the present invention.
- FIG. 1 is a block diagram showing a configuration of a distributed data storage system according to the first embodiment of the present invention.
- one management node 4 and two or more storage nodes 5 having at least the same performance are connected via a network so that they can communicate with each other.
- At least one client device 1 is connected to a plurality of storage nodes 5 and a plurality of storage nodes 6 added later via a network 2 so as to be able to communicate with each other.
- the client device 1 is a computer operated by a user. When there is a request for new writing, the user operates the client device 1 to transmit a data access request for accessing the storage node 5 of the distributed data storage system 3.
- the client apparatus 1 uses a predetermined arithmetic expression in order to calculate the data storage destination. That is, the client device 1 assigns an ID to data to be written, uses the ID as arithmetic input data, and calculates a value indicating a data storage destination.
- the client device 1 collates the calculated value with the storage information included in the system configuration information acquired from the management node 4, and sets the data storage destination. Further, the client device 1 reads and writes data by transmitting the ID and the data to be written together with the write command to the storage node 5 that is the data storage destination.
- the management node 4 is a terminal device operated by an administrator of the distributed data storage system 3.
- An administrator of the distributed data storage system 3 can operate the management node 4 to access the storage nodes 5 and 6, set system configuration information, and perform various settings necessary for operation.
- FIG. 2 is a block diagram showing the configuration of the management node according to the first embodiment of the present invention. As shown in FIG. 2, the input / output unit 7, the setting unit 10, the determination unit 12, the transmission unit 13, and the storage unit 14 are connected to the management node 4.
- the input / output unit 7 is connected to the keyboard 8 and the mouse 9 and transfers signals sent from the keyboard 8 and the mouse 9 to each unit via the bus.
- the setting unit 10 is connected to the monitor 11. When a system configuration change occurs due to the installation, addition, or failure of storage nodes having different I / O performance in the distributed data storage system 3, the setting unit 10 displays the system configuration information on the monitor screen.
- the system configuration information includes the logical data capacity, storage node IP address information, storage node availability, and a calculation result based on a predetermined arithmetic expression. Storage information that associates a value with a data storage destination ”.
- the setting unit 10 sets the latest system configuration information reflecting the system configuration change based on the installation, addition, failure, or the like of the storage node.
- the setting unit 10 divides the storage area of the installed or additional storage node into a first storage area and a second storage area, and sets the logical capacity of each area.
- the determination unit 12 determines system configuration information reflecting the configuration change of the distributed data storage system based on the information set by the setting unit 10.
- the transmission unit 13 transmits the information set by the setting unit 10 and the system configuration information determined by the determination unit 12 to the client device 1 and / or each storage node via the network 2.
- the storage unit 14 stores information set by the setting unit 10 and system configuration information determined by the determination unit 12.
- Consistent Hashing is used as an arithmetic expression for the client apparatus 1 to determine the data placement destination.
- the management node 4 assigns virtual nodes according to the storage capacity installed in each node. Further, the number of virtual nodes and the virtual node number are assigned according to the logical data capacity set for each node.
- Reference node 1 Virtual node 1
- Reference node 2 Virtual node 2
- Node 1 with double logical capacity 1 Virtual nodes 3, 4
- the number of each virtual node is a result (value) calculated by an arithmetic expression, and the correspondence relationship of the virtual node in each node is stored information in the storage unit 14.
- Pieces of information are set by the system operation manager, and are set information set in advance by the management node 4.
- the management node 4 manages the storage nodes 5 and 6 via the network 2 based on the above configuration.
- management node 4 may be configured by any one storage node 5.
- FIG. 3 is a block diagram showing the configuration of the storage node 5.
- the storage node 5 includes a data input / output management unit 15, a data storage destination management unit 16, a storage usage management unit 17, and a storage unit 18 having a storage device 19. Each unit is connected to the internal bus of the storage node 5.
- each storage node 5 has a storage device 19 connected thereto.
- the storage node 5 manages the data stored in the storage device 19 and provides the managed data to the client device 1 via the network 2.
- the storage node 5 manages data with redundancy. That is, data having the same content is managed by at least two storage nodes.
- the storage device 19 is composed of a single hard disk (HDD) without using a disk array.
- a storage device may be composed of a plurality of devices with different performances (for example, a combination of SSD and HDD).
- the storage device 19 may be a RAID (Redundant Array of Inexpensive Disks) system using a plurality of built-in HDDs, or a disk array may be configured using a technique other than RAID.
- the storage device provides a disk management service for a single hard disk (HDD).
- the data input / output management unit 15 is in charge of input / output of data transmitted / received via the network 2.
- the client device 1 When a system configuration change occurs in the distributed data storage system 3 due to an addition or failure of a storage node, the client device 1 does not hold system configuration information reflecting the system configuration change generated in the management node 4. Further, it is conceivable that the client device 1 transmits a write command to an incorrect node.
- the data input / output management unit 15 performs a calculation based on an arithmetic algorithm with the data ID included in the write command transmitted to the own node as an input, and the value of the calculation result and the system configuration from the management node 4
- the stored information included in the information is collated to determine whether the data access request is a request for the own node.
- the data input / output management unit 15 determines that the data access request is a request for the own node, and according to the data access request, the data ID and the data read / write The command is transmitted to the data storage destination management unit 16.
- the data input / output management unit 15 determines that the request is not for the node itself, and returns an error indicating that the data storage destination is not the client device 1.
- the client device 1 When acquiring the error, the client device 1 requests the management node 4 to transmit the latest system configuration information and assigns an ID to the data to be written. Further, in order to calculate the data storage destination, the client device 1 uses a predetermined arithmetic expression and calculates a value indicating the data storage destination using the ID as input data.
- the client device 1 collates the value calculated by the arithmetic expression with the storage information included in the latest system configuration information transmitted from the management node 4, and resets the data storage destination.
- the client device 1 reads and writes data by retransmitting the ID and the data to be written together with the write command to the storage node 5 that is the data storage destination that has been reset.
- the data storage destination management unit 16 manages a table that associates the physical address that is the data storage destination of the storage device 19 of the own node with the ID of the data to be written requested by the data input / output management unit 15. Yes.
- the data storage destination management unit 16 reads / writes data from the storage device 19 in accordance with the data ID specified by the data input / output management unit 15 and the data read / write command, and is given to the data input / output management unit 15. Returns the response of the command.
- the contents of the above-described table are general data management tables in which data IDs, storage device physical addresses, and access frequency information are linked.
- the data storage destination management unit 16 can extract the physical address and access frequency information of the storage device with the data ID as an input.
- the data storage destination management unit 16 registers the request as access frequency information of the data ID in the table.
- the storage usage management unit 17 manages free addresses of the storage device 19.
- the data storage destination management unit 16 acquires a free address from the storage usage management unit 17 and stores the storage device 19. Write to.
- the address at which data was written to the storage device 19 is managed as an address in use by the storage usage management unit 17.
- the storage usage management unit 17 sets all the storage devices 19 based on the logical data capacity included in the system configuration information transmitted from the management node 4 when the storage node 5 is incorporated into the distributed data storage system 3. Are classified into two types: an address for the first storage area and an address for the second storage area. In this case, the storage usage management unit 17 classifies the addresses so that the logical data capacity satisfies the capacity of the first storage area and the capacity of the second storage area.
- the storage device 19 is configured by a general storage device such as an HDD or an SSD, but may be configured by combining a plurality of different devices. When a plurality of different devices are combined, the storage usage management unit 17 allocates a device with higher I / O performance as the first storage area.
- the newly introduced storage node 6 may be mounted using the same hardware as the existing storage node 5.
- FIGS. 4 to 6 are flowcharts showing data write processing from the client apparatus 1 to the distributed data storage system 3 when a storage node 6 having different I / O performance is newly introduced. The processing shown in FIGS. 4 to 6 will be described below with reference to flowcharts.
- Step S401 As shown in FIG. 1, as the I / O performance value of the existing storage node 5, the distributed storage system 3 having a logical capacity of 10 is added to the storage node 6 having an I / O performance value of 20. Is newly introduced. In this case, the management node 4 sets the system configuration information reflecting the system configuration change based on the addition of the storage nodes 6 having different performances.
- Step S402 The management node 4 divides the storage device managed by the newly introduced storage node 6 into a first storage area and a second storage area, and sets the respective capacities based on the following conditions: Assign.
- the following conditions are rules for setting the capacity allocated to each node.
- the management node 4 When setting the capacity allocated to each node, the management node 4 is set to match the I / O performance value of the node constituting the system as a part of the information included in the system configuration information managed by the management node 4. . That is, the management node 4 sets the capacity based on the following rules.
- Condition 1) The management node 4 determines the I / O performance value of the reference storage node 5.
- the management node 4 determines the logic of the first storage of the storage node 6 with different I / O performance.
- the capacity is set to N times the reference node (N is an integer greater than 0).
- the management node 4 determines the logical capacity of the first storage of the storage node 6 with different I / O performance. Is set to twice the logical capacity of the reference node to match the I / O performance of the storage node 6. As a result, the problem that the performance of nodes with different I / O performance becomes a bottleneck can be solved, so that the performance of the entire system can be maintained.
- Condition 2 When the capacity of the storage device in the storage node 6 having different I / O performance is equal to or more than N times that of the storage device in the reference storage node 5 in the condition 1), the management node 4 has N in the storage node 6 The remaining capacity exceeding the double is set as the second storage area, and the second storage area is not set in the reference storage node 5 (not the data relocation destination).
- the management node 4 has the remaining capacity exceeding twice in the storage node 6. Is set as the second storage area.
- the storage node 6 can hold twice the capacity in the first storage area in accordance with the I / O performance, so the performance of the storage node 6 with different I / O performance becomes a bottleneck. Can be solved, and the performance of the entire system can be maintained.
- Condition 3 When the second storage area is set in the reference storage node 5 in the conditions 1) and 2), the capacity that is N times the capacity allocated to the first storage area of the reference storage node is I / O. Allocation is performed as the capacity of the first storage area of the storage node 6 having different performance.
- the logical capacity of the first storage area of the storage node 6 is the first storage area of the reference storage node 5. Is set to twice the capacity allocated to.
- the storage node 6 can hold twice the capacity in the first storage area in accordance with the I / O performance, so the performance of the storage node 6 with different I / O performance becomes a bottleneck. Can be solved, and the performance of the entire system can be maintained.
- Condition 4) In the condition 1), when the capacity of the storage device in the storage node 6 with different I / O performance is N times or less than the storage device of the reference storage node 5, the management node 4 has different I / O performance. All the storage devices in the storage node 6 are allocated as the first storage area with the upper limit being the capacity of the storage device.
- the I / O performance of the storage node 6 is twice the I / O performance of the reference storage node 5 as compared to the reference node, and the capacity of the storage device in the storage node 6 having different I / O performance is the reference storage node 5
- the management node 4 allocates all the storage devices in the storage node 6 having different I / O performances as the first storage area with the upper limit.
- Condition 5 In the conditions 1) and 4), 1 / N capacity of the capacity of the first storage area in the storage node 6 having different I / O performance is allocated as the capacity of the first storage area of the reference storage node.
- the I / O performance of the storage node 6 is twice the I / O performance of the reference storage node 5, and the storage device capacity in the storage node 6 with different I / O performance is the storage device of the reference storage node 5. It is assumed that it is 2 times or less.
- the management node 4 assigns a capacity of 1 ⁇ 2 of the capacity of the first storage area in the storage node 6 having different I / O performance as the capacity of the first storage area of the reference storage node 5.
- condition 1) the management node 4 can adjust the logical capacity to the I / O performance, and can solve the problem that the performance of nodes with different I / O performance becomes a bottleneck, The overall system performance can be maintained.
- Condition 6) When there are three or more types of I / O performance and storage nodes 6 with different storage logical capacities, the node with the lowest I / O performance is set as the reference storage node. Based on conditions 1) to 5), The capacity of the first storage area and the capacity of the second storage area are set.
- the node with the lowest I / O performance is set as the reference storage node. Therefore, based on the conditions 1) to 5), the logical capacity of the first storage area of the storage node with the lowest capacity is used as a reference. It can be easily adjusted to the I / O performance of the storage node.
- Condition 7 When the storage nodes 5 and 6 are combinations of storage nodes that do not meet the conditions 1) to 6), the storage node having the smallest storage capacity is set as the reference storage node.
- the management node 4 can easily solve the problem that the storage node becomes a bottleneck by setting the I / O performance of the storage node according to the capacity of the reference storage node.
- the management node 4 selects only the device with the highest I / O performance.
- the capacity of the first storage area and the logical capacity of the second storage area are set according to the conditions 1) to 6) based on the I / O performance and the storage logical capacity when used.
- the management node 4 is set based on the condition 1). That is, the management node 4 sets the logical capacity of the first storage area of the introduced storage node 6 to 20 which is twice the capacity of the reference storage node 5.
- the logical capacity of the newly introduced storage node 6 can be matched with the I / O performance, so that there is no difference between the logical capacity of the storage node 6 and the I / O performance.
- Step S403 The management node 4 transmits the changed system configuration information and setting information such as the logical capacity of the first storage area to the client device 1 and all the storage nodes 5 and 6. Alternatively, the management node 4 may transmit the information when there is a request for transmission of changed system configuration information or setting information from the client device 1 or the storage nodes 5 and 6.
- Step S404 When there is a request for new writing from the user, the client apparatus 1 assigns an ID to the data to be written, and uses the ID as input data, a predetermined arithmetic expression, and the management node 4 Based on the acquired system configuration information, the storage node 5 as the data storage destination is set.
- Step S405 The client apparatus 1 transmits the ID and the data to be written together with the write command to the storage node 5 that is the storage destination of the data set in Step S404.
- Step S406 The data input / output management unit 15 of the storage node 5 calculates the data ID included in the write command transmitted to the own node as an input by an arithmetic algorithm. Further, the data input / output management unit 15 collates the calculated value with the storage information included in the system configuration information from the management node 4, and determines whether the data access request is a request for the own node.
- the data input / output management unit 12 determines that the own node is the storage destination, and the data input / output management unit 15 writes the write request transmitted from the client device 1. Is transferred to the data storage destination management unit 16.
- Step S407 The data storage destination management unit 16 checks whether the data ID included in the write request exists in the table in which the data ID and the storage device 19 are associated.
- Step S408 The data storage destination management unit 16 acquires an address associated with the table.
- Step S409 The data storage destination management unit 16 writes the data included in the write request to an appropriate address of the storage device 19 based on the acquired address.
- Step S410 The data storage destination management unit 16 transfers the completion of writing to the data input / output management unit 15. In response to the completion of writing, the data input / output management unit 15 notifies the client device 1 that the write request has been completed.
- Step S ⁇ b> 4111 the data storage destination management unit 16 acquires a free address from the storage usage management unit 17 in the first storage area. Next, the data storage destination management unit 16 associates the ID included in the data write request with the empty address, registers it as a new entry on the table, and completes the writing.
- Step S412 When the usage amount of the first storage area becomes a certain amount or more, and the usage amount of the storage device secured as the first storage area is equal to or more than the predetermined usage amount, In other words, when the number of free addresses is equal to or less than a predetermined number, the storage usage management unit 17 raises an alarm of excess usage to the data storage destination management unit 16.
- the data storage destination management unit 16 refers to the access frequency information of each ID managed on the table in which the ID and the data storage destination address to the storage device 19 are linked with an alarm as a trigger. Next, the data storage destination management unit 16 extracts ID groups for a predetermined fixed amount of data from IDs with the lowest access frequency.
- Step S414 The data storage destination management unit 16 performs a cyclic bit shift by a predetermined number of binary data constituting the extracted ID. Thereafter, the data storage destination management unit 16 uses the shifted ID data as input data, system configuration information shared among the client device 1, the storage nodes 5, 6 and the management node 4, and an arithmetic data storage destination. Using this setting method, a new data storage destination is set. The storage destination is the ID data re-storage destination.
- Step S415 ⁇ When the re-storage destination is the local node> (Step S415) As shown in FIG. 5, the data storage destination management unit 16 instructs the storage usage management unit 17 to change the address associated with the ID from the first storage area to the second storage area. Notify that the assignment has changed.
- Step S416) Upon receipt of the notification, the storage usage management unit 17 registers the corresponding address as the address of the second storage area, and from the addresses assigned to the second storage area, the free address Is selected and registered as the address of the first storage area.
- the data storage destination management unit 16 may rearrange data from the first storage area to the second storage area.
- Step S41-7 The data storage destination management unit 16 writes the data included in the write request to the address registered as the address of the first storage area.
- Step S4128 The data storage destination management unit 16 transfers the completion of writing to the data input / output management unit 15. In response to the completion of writing, the data input / output management unit 15 notifies the client device 1 that the write request has been completed.
- Step S419) As shown in FIG. 6, the data storage destination management unit 16 reads the data corresponding to the ID from the storage device 19, and the address corresponding to the ID is released (becomes a free address). Is notified to the storage usage management unit 17. Furthermore, the data storage destination management unit 16 replaces the address information with an identifier indicating that it has been re-stored in a table that manages the ID and the address information of the storage device 19 in association with each other.
- Step S420 the data storage destination management unit 16 sets the data read from the storage device 19 and the ID after re-stored by cyclic bit shift, as the storage destination via the data input / output management unit 12. Transfer to the storage node 6 together with the re-store command.
- Step S421 The data input / output management unit 15 of the storage node 6 that has become the re-storage destination receives the re-storage request, and then transfers the request to the data storage destination management unit 16.
- the data storage destination management unit 16 acquires a free address from the addresses reserved as the second storage area from the storage usage management unit 17 and registers it as a new entry in the table that associates the ID with the address. After that, the data is stored in the free address.
- the data storage destination management unit 16 may acquire a free address in the first storage area and store the data at the address.
- Step S422 The notification of completion of the processing in Step S421 is transferred to the storage node 5 that is the storage source via the data input / output management unit 15. Restorage source storage node 5 receives the write completion notification and notifies client device 1 that the write request has been completed.
- the management node 4 when configured with storage node groups having different I / O performances, has the logical capacity and the I / O performance of each storage node.
- the logical capacity of the storage node can be set based on the number of data access I / Os in accordance with the performance difference.
- FIG. 7 is a block diagram showing the configuration of the storage node management apparatus 700 according to the second embodiment of the present invention. As illustrated in FIG. 7, the storage node management apparatus 700 includes a setting unit 701, a determination unit 702, and a transmission unit 703.
- the storage node management device 700 is connected to a plurality of storage nodes 5 and 6 via the network 2.
- An administrator of the distributed data storage system 3 operates the storage node management device 700. Based on the operation, the storage node management device 700 accesses the storage nodes 5 and 6 having different I / O performances, sets system configuration information, and performs various settings necessary for operation.
- the setting unit 701 displays the system configuration information on the monitor screen.
- the system configuration information is calculated based on “logical data capacity”, “storage node IP address information”, “presence / absence of storage node operation”, and “calculation result based on a predetermined arithmetic expression”.
- Storage information for associating a value with a data storage destination ”.
- the setting unit 10 sets the latest system configuration information that reflects the system configuration change that has occurred, based on the installation, addition, or failure of the storage node.
- the setting unit 10 divides the storage area of the installed or additional storage node into a first storage area and a second storage area, and sets the logical capacity of each area.
- the determination unit 702 determines system configuration information reflecting the configuration change of the distributed data storage system 3 based on the information set by the setting unit 701.
- the transmission unit 703 transmits the information set by the setting unit 701 and the system configuration information determined by the determination unit 702 to the client device 1 or each of the storage nodes 5 and 6 via the network 2.
- the storage node management device 700 uses the storage node logical capacity and I / O performance.
- the logical capacity of the storage node can be set based on the number of data access I / Os in accordance with the respective performance differences.
- the data storage capacity of the storage node is set to the number of data access I / Os according to the performance difference between the logical capacity and the I / O performance of the storage node. It becomes possible to set based on.
- the performance difference between the logical capacity and I / O performance of the storage nodes constituting the distributed data storage system can be eliminated or minimized, and the H / W performance of each storage node can be made as close as possible to the maximum. Resources can be allocated. As a result, the overall performance of the distributed data storage system can be improved in both storage capacity and I / O performance.
- the processing functions described with reference to the flowcharts can be realized by a computer.
- a program describing the processing contents of the functions that the storage nodes 5 and 6, the management node 4, and the storage node management apparatus 700 should have is provided.
- FIG. 8 is a circuit block diagram of a computer device of the management node 4 as the first embodiment of the present invention or the storage node management device 700 as the second embodiment.
- the processing function of the management node 4 as the first embodiment of the present invention or the storage node management device 700 as the second embodiment is a ROM 803 (Read Only Memory) or a storage device (HDD).
- the CPU 801 Central Processing Unit
- the RAM 802 Random Access Memory
- a program describing contents for realizing the processing function can be recorded on a computer-readable recording medium.
- the computer-readable recording medium include a magnetic recording device, an optical disk, a magneto-optical recording medium, and a semiconductor memory.
- Examples of the magnetic recording device include an HDD, a flexible disk (FD), and a magnetic tape (MT).
- Optical discs include DVD (Digital Versatile Disc), DVD-RAM, CD-ROM (Compact Disc-Read Only Memory), CD-R (Recordable) / RW (Rewritable), and the like.
- Magneto-optical recording media include MO (Magneto-Optical disk).
- a portable recording medium such as a DVD or CD-ROM in which the program is recorded is sold. It is also possible to store the program in a server computer and transfer the program from the server computer to another computer via a network.
- the computer that executes the program stores, for example, the program recorded on the portable recording medium or the program transferred from the server computer in its own storage device.
- the computer reads the program from its own storage device and executes processing according to the program.
- the computer can also read the program directly from the portable recording medium and execute processing according to the program.
- the computer can sequentially execute processing according to the received program.
- the present invention can be applied to a case where storage nodes having different capacities and I / O performance are mixedly loaded in the distributed data storage system from the beginning.
- the present invention may be a combination of any two or more configurations (features) of the above-described embodiments.
- At least one reference storage node is set from the plurality of storage nodes having different I / O performance, and the I / O of the reference storage node is set.
- the storage areas of the storage nodes other than the reference storage node are divided into the first storage area and the second storage area, and the logic of the first and second storage areas is divided.
- Setting means for setting the capacity to match the I / O performance of storage nodes other than the reference storage node; Determining means for determining system configuration information reflecting a configuration change of the distributed data storage system based on the information set by the setting means;
- a storage node management apparatus comprising: transmission means for transmitting the system configuration information determined by the determination means to the plurality of storage nodes having different I / O performances.
- the setting means sets a capacity N times the logical capacity set in the first storage area of the reference storage node to the reference storage node.
- the storage node management apparatus according to any one of appendices 1 to 3, wherein the logical capacity of the first storage area of the other storage node is set.
- the setting means sets the storage node having the lowest I / O performance as a reference storage node, and sets the I / O of all storage nodes.
- the storage management according to any one of appendices 1 to 4, wherein the logical capacity of the first storage area and the logical capacity of the second storage area are set so that the O performance and the storage capacity approach the maximum apparatus.
- the setting means sets the node having the smallest storage capacity as the reference storage node, and sets the I / O performance and storage capacity of all the storage nodes to the maximum.
- the storage management device according to any one of appendices 1 to 5, wherein a logical capacity of the storage area and a logical capacity of the second storage area are set.
- At least one reference storage node is set from the plurality of storage nodes having different I / O performances.
- the storage area of the storage node other than the reference storage node is divided into a first storage area and a second storage area, and the first and second storage areas
- the storage node logical capacity setting method is set so as to match the I / O performance of the storage nodes other than the reference storage node.
- Storage means having a storage device divided into a first storage area and a second storage area in the storage management device according to any one of appendices 1 to 6;
- data input / output management means for determining whether the access request is for the own node;
- a data storage management means for reading and writing data in accordance with a command to be instructed; Based on the logical capacity included in the system configuration information acquired from the storage management device according to any one of appendices 1 to 6, the storage is performed so that the first storage area and the second storage area are filled with capacity.
- a storage node comprising storage usage management means for classifying device addresses.
- Appendix 14 A distributed data storage system comprising: the storage management device according to any one of appendices 1 to 6; and the storage node according to appendix 13.
Abstract
Description
前記設定部により設定した情報に基づいて、前記分散データストレージシステムの構成変更を反映したシステム構成情報を決定する決定手段と、
前記決定部により決定したシステム構成情報を前記複数のI/O性能の異なるストレージノードに送信する送信手段とを備える。
前記格納部に格納するデータにアクセス要求がある場合、自ノード向けのアクセス要求であるかどうかを判定するデータ入出力管理手段と、
書き込み対象となるデータのID、データ格納先となる物理アドレス、アクセス頻度情報の3つの情報を紐づけたデータ管理テーブルを管理し、前記データ入出力管理部によって指定されたデータのIDとデータの読み書きの命令に応じて、データの読み書きを行うデータ格納管理手段と、
前記ストレージ管理装置から取得するシステム構成情報に含まれる論理容量に基づき、前記第1のストレージ領域と第2のストレージ領域に容量を満たすように前記ストレージデバイスのアドレスを分類するストレージ使用量管理手段とを備える。
[第1の実施形態]
図1は、本発明の第1の実施形態に係る分散データストレージシステムの構成を示すブロック図である。図1に示すように、分散データストレージシステム3は、1つの管理ノード4および少なくとも性能は同一の2台以上のストレージノード5がネットワークを介して相互通信可能に接続されている。
例えば、次のように割り当てられる。
基準ノード2:バーチャルノード2
論理容量が2倍のノード1:バーチャルノード3、4
論理容量が2倍のノード2:バーチャルノード5、6
各バーチャルノードの番号が、算術式により算出される結果(値)であり、上記の各ノードにおけるバーチャルノードの対応関係が記憶部14の格納情報となる。
条件1) 管理ノード4は、基準となるストレージノード5のI/O性能値を決める。次に、I/O性能の異なるストレージノード6のI/O性能が基準ストレージノード5のN倍である場合、管理ノード4は、I/O性能の異なるストレージノード6の第1のストレージの論理容量を基準ノードのN倍に設定する(Nは、0より大きい整数である)。
条件2) 条件1)において、I/O性能の異なるストレージノード6内のストレージデバイスの容量が、基準ストレージノード5のストレージデバイスのN倍以上ある場合、管理ノード4は、ストレージノード6内にN倍を超える残りの容量を、第2のストレージ領域として設定し、基準ストレージノード5には第2のストレージ領域を設定しない(データ再配置先とならない)。
条件3) 条件1)及び条件2)において、基準ストレージノード5に第2のストレージ領域を設定する場合、基準ストレージノードの第1のストレージ領域に割り当てた容量のN倍の容量を、I/O性能の異なるストレージノード6の第1のストレージ領域の容量として割り当てる。
条件4) 条件1)において、I/O性能の異なるストレージノード6内のストレージデバイスの容量が、基準ストレージノード5のストレージデバイスのN倍以下の場合、管理ノード4は、I/O性能の異なるストレージノード6内のストレージデバイスの容量を上限として第1のストレージ領域としてすべて割り当てる。
条件5) 条件1)及び4)において、I/O性能の異なるストレージノード6における第1のストレージ領域の容量の1/Nの容量を、基準ストレージノードの第1のストレージ領域の容量として割り当てる。
条件6) 3種類以上のI/O性能、ストレージ論理容量の異なるストレージノード6が存在した場合、最もI/O性能の低いノードを基準ストレージノードとして設定し、条件1)~5)に基づき、第1のストレージ領域の容量、第2のストレージ領域の容量を設定する。
条件7) ストレージノード5、6が条件1)~6)に合致しないストレージノードの組み合わせである場合、最も少ないストレージ容量となるストレージノードを基準ストレージノードとする。
(ステップS408)データ格納先管理部16は、テーブルに紐づけられているアドレスを取得する。
(ステップS411)図5に示すように、データ格納先管理部16は、ストレージ使用量管理部17から、第1のストレージ領域の中から空きアドレスを取得する。次に、データ格納先管理部16は、データ書き込み要求に含まれるIDと空きアドレスを紐づけて、テーブル上に新たなエントリとして登録して書き込みを完了する。
(ステップS415)図5に示すように、データ格納先管理部16は、ストレージ使用量管理部17に対して、IDに紐づけられたアドレスが、第1のストレージ領域から第2のストレージ領域に割り当てが変更になったことを通知する。
(ステップS419)図6に示すように、データ格納先管理部16は、ストレージデバイス19から前記IDに該当するデータを読み出し、前記IDに該当するアドレスが解放された(空きアドレスとなった)ことをストレージ使用量管理部17に通知する。さらに、データ格納先管理部16は、IDとストレージデバイス19のアドレス情報を紐づけて管理しているテーブルにおいて、アドレス情報を、再格納されたことを示す識別子に置き換える。
[第2の実施形態]
次に、本発明の第2の実施形態について図面を参照して詳細に説明する。なお、本実施形態の説明において、本発明の第2の実施形態に係るストレージノード管理装置700の構成と本発明の第1の実施形態に係る管理ノードの構成の以外に、本発明の第1の実施形態と同様なシステム構成や同様に動作するステップを有するので、それにおける詳細な説明を省略する。
I/O性能の異なる複数のストレージノードを含む分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定し、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定する設定手段と、
前記設定手段により設定した情報に基づいて、前記分散データストレージシステムの構成変更を反映したシステム構成情報を決定する決定手段と、
前記決定手段により決定したシステム構成情報を前記複数のI/O性能の異なるストレージノードに送信する送信手段とを備えるストレージノード管理装置。
前記設定手段は、前記基準ストレージノード以外のストレージノードのI/O性能が前記基準ストレージノードのI/O性能のN倍である場合、前記基準ストレージノード以外のストレージノードの第1のストレージ論理容量を前記基準ストレージノードの論理容量のN倍に設定することを特徴とする付記1に記載のストレージノード管理装置。
前記設定手段は、前記基準ストレージノード以外のストレージノードの論理容量が、前記基準ストレージノードの論理容量のN倍以上である場合、N倍を超える残りの容量を、前記基準ストレージノード以外のストレージノードの第2のストレージ領域に設定することを特徴とする付記1または2に記載のストレージノード管理装置。
前記設定手段は、前記基準ストレージノードに第1及び第2のストレージ領域が設定される場合、前記基準ストレージノードの第1のストレージ領域に設定した論理容量のN倍の容量を、前記基準ストレージノード以外のストレージノードの第1のストレージ領域の論理容量に設定することを特徴とする付記1乃至3のいずれか1つに記載のストレージノード管理装置。
前記設定手段は、3種類以上のI/O性能、ストレージ容量の異なるストレージノードが存在した場合は、最もI/O性能の低いストレージノードを基準ストレージノードとして設定し、全てのストレージノードのI/O性能及び格納容量が最大に近づくように第1のストレージ領域の論理容量、第2のストレージ領域の論理容量を設定することを特徴とする付記1乃至4のいずれか1つに記載のストレージ管理装置。
前記設定手段は、性能の差異が大きいストレージノードの組み合わせの場合、最も少ないストレージ容量となるノードを基準ストレージノードとし、全てのストレージノードのI/O性能及び格納容量が最大に近づくように第1のストレージ領域の論理容量、第2のストレージ領域の論理容量を設定することを特徴とする付記1乃至5のいずれか1つに記載のストレージ管理装置。
I/O性能の異なる複数のストレージノードにより基づいて、分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定し、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定するストレージノード論理容量設定方法。
前記基準ストレージノード以外のストレージノードのI/O性能が前記基準ストレージノードのI/O性能のN倍である場合、前記基準ストレージノード以外のストレージノードの第1のストレージ論理容量を前記基準ストレージノードの論理容量のN倍に設定することを特徴とする付記7に記載のストレージノード論理容量設定方法。
前記基準ストレージノード以外のストレージノードのストレージ論理容量が、前記基準ストレージノードのストレージ容量のN倍以上である場合、N倍を超える残りの容量を、前記基準ストレージノード以外のストレージノードの第2のストレージ領域に設定することを特徴とする付記7又は8に記載のストレージノード論理容量設定方法。
前記基準ストレージノードに第1及び第2のストレージ領域が設定される場合、前記基準ストレージノードの第1のストレージ領域に設定した容量のN倍の容量を、前記基準ストレージノード以外のストレージノードの第1のストレージ領域の容量に設定することを特徴とする付記7乃至9のいずれか1つに記載のストレージノード論理容量設定方法。
I/O性能の異なる複数のストレージノードを含む分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定する処理と、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定する処理とをコンピュータに実行させるプログラム。
I/O性能の異なる複数のストレージノードを含む分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定する処理と、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定する処理とをコンピュータに実行させるプログラムが記録された記録媒体。
前記付記1乃至6のいずれか1つに記載のストレージ管理装置で第1のストレージ領域と第2のストレージ領域に分割されるストレージデバイスを有する格納手段と、
前記格納手段に格納するデータにアクセス要求がある場合、自ノード向けのアクセス要求であるかどうかを判定するデータ入出力管理手段と、
書き込み対象となるデータのIDと、データ格納先となる物理アドレスと、アクセス頻度情報とを紐づけた情報を管理すると共に、前記データ入出力管理手段によって指定されたデータのIDとデータの読み書きを指示する命令に応じて、データの読み書きを行うデータ格納管理手段と、
前記付記1乃至6のいずれか1つに記載のストレージ管理装置から取得するシステム構成情報に含まれる論理容量に基づき、前記第1のストレージ領域と第2のストレージ領域に容量を満たすように前記ストレージデバイスのアドレスを分類するストレージ使用量管理手段とを備えるストレージノード。
付記1乃至6のいずれか1つに記載のストレージ管理装置と、付記13に記載のストレージノードとを備える分散データストレージシステム。
2 ネットワーク
3 分散データストレージシステム
4 管理ノード
5 ストレージノード
6 ストレージノード
7 入出力部
8 キーボード
9 マウス
10 設定部
11 モニタ
12 決定部
13 送信部
14 記憶部
15 データ入出力管理部
16 データ格納先管理部
17 ストレージ使用量管理部
18 格納部
19 ストレージデバイス
700 ストレージノード管理装置
701 設定部
702 決定部
703 送信部
801 CPU
802 RAM
803 ROM
804 記憶装置
805 外部機器接続インタフェース
806 ネットワークインタフェース
Claims (10)
- I/O性能の異なる複数のストレージノードを含む分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定し、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定する設定手段と、
前記設定手段により設定した情報に基づいて、前記分散データストレージシステムの構成変更を反映したシステム構成情報を決定する決定手段と、
前記決定手段により決定したシステム構成情報を前記複数のI/O性能の異なるストレージノードに送信する送信手段とを備えるストレージノード管理装置。 - 前記設定手段は、前記基準ストレージノード以外のストレージノードのI/O性能が前記基準ストレージノードのI/O性能のN倍である場合、前記基準ストレージノード以外のストレージノードの第1のストレージ論理容量を前記基準ストレージノードの論理容量のN倍に設定することを特徴とする請求項1に記載のストレージノード管理装置。
- 前記設定手段は、前記基準ストレージノード以外のストレージノードの論理容量が、前記基準ストレージノードの論理容量のN倍以上である場合、N倍を超える残りの容量を、前記基準ストレージノード以外のストレージノードの第2のストレージ領域に設定することを特徴とする請求項1または2に記載のストレージノード管理装置。
- 前記設定手段は、前記基準ストレージノードに第1及び第2のストレージ領域が設定される場合、前記基準ストレージノードの第1のストレージ領域に設定した論理容量のN倍の容量を、前記基準ストレージノード以外のストレージノードの第1のストレージ領域の論理容量に設定することを特徴とする請求項1乃至3のいずれか1つに記載のストレージノード管理装置。
- I/O性能の異なる複数のストレージノードによって、分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定し、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定するストレージノード論理容量設定方法。
- 前記基準ストレージノード以外のストレージノードのI/O性能が前記基準ストレージノードのI/O性能のN倍である場合、前記基準ストレージノード以外のストレージノードの第1のストレージ論理容量を前記基準ストレージノードの論理容量のN倍に設定することを特徴とする請求項5に記載のストレージノード論理容量設定方法。
- I/O性能の異なる複数のストレージノードを含む分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定する処理と、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定する処理とをコンピュータに実行させるプログラム。
- I/O性能の異なる複数のストレージノードを含む分散データストレージシステムを構成する際、前記複数のI/O性能の異なるストレージノードから少なくとも1つの基準ストレージノードを設定する処理と、前記基準ストレージノードのI/O性能と論理容量とを参照して、前記基準ストレージノード以外のストレージノードのストレージ領域を第1のストレージ領域と第2のストレージ領域に分割すると共に、それら第1及び第2のストレージ領域の論理容量を前記基準ストレージノード以外のストレージノードのI/O性能に合わせるように設定する処理とをコンピュータに実行させるプログラムが記録された記録媒体。
- 前記請求項1乃至4のいずれか1つに記載のストレージ管理装置で第1のストレージ領域と第2のストレージ領域に分割されるストレージデバイスを有する格納手段と、
前記格納手段に格納するデータにアクセス要求がある場合、自ノード向けのアクセス要求であるかどうかを判定するデータ入出力管理手段と、
書き込み対象となるデータのIDと、データ格納先となる物理アドレスと、アクセス頻度情報とを紐づけた情報を管理すると共に、前記データ入出力管理手段によって指定されたデータのIDとデータの読み書きを指示する命令に応じて、データの読み書きを行うデータ格納管理手段と、
前記請求項1乃至4のいずれか1つに記載のストレージ管理装置から取得するシステム構成情報に含まれる論理容量に基づき、前記第1のストレージ領域と第2のストレージ領域に容量を満たすように前記ストレージデバイスのアドレスを分類するストレージ使用量管理手段とを備えるストレージノード。 - 請求項1乃至4のいずれか1つに記載のストレージ管理装置と、請求項9に記載のストレージノードとを備える分散データストレージシステム。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/781,088 US10708355B2 (en) | 2013-05-20 | 2014-05-15 | Storage node, storage node administration device, storage node logical capacity setting method, program, recording medium, and distributed data storage system |
JP2015518069A JP6222227B2 (ja) | 2013-05-20 | 2014-05-15 | ストレージノード、ストレージノード管理装置、ストレージノード論理容量設定方法、プログラム、記録媒体および分散データストレージシステム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2013106077 | 2013-05-20 | ||
JP2013-106077 | 2013-05-20 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014188682A1 true WO2014188682A1 (ja) | 2014-11-27 |
Family
ID=51933251
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2014/002562 WO2014188682A1 (ja) | 2013-05-20 | 2014-05-15 | ストレージノード、ストレージノード管理装置、ストレージノード論理容量設定方法、プログラム、記録媒体および分散データストレージシステム |
Country Status (3)
Country | Link |
---|---|
US (1) | US10708355B2 (ja) |
JP (1) | JP6222227B2 (ja) |
WO (1) | WO2014188682A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10860541B2 (en) | 2016-04-11 | 2020-12-08 | Johnson Controls Fire Protection LP | Fire detection system with distributed file system |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9367243B1 (en) | 2014-06-04 | 2016-06-14 | Pure Storage, Inc. | Scalable non-uniform storage sizes |
US9916275B2 (en) * | 2015-03-09 | 2018-03-13 | International Business Machines Corporation | Preventing input/output (I/O) traffic overloading of an interconnect channel in a distributed data storage system |
WO2017119091A1 (ja) * | 2016-01-07 | 2017-07-13 | 株式会社日立製作所 | 分散型ストレージシステム、データ格納方法、およびソフトウェアプログラム |
CN111506254B (zh) * | 2019-01-31 | 2023-04-14 | 阿里巴巴集团控股有限公司 | 分布式存储系统及其管理方法、装置 |
JP6857673B2 (ja) * | 2019-02-14 | 2021-04-14 | 株式会社日立製作所 | マルチストレージノードシステム、マルチストレージノードシステムの容量管理方法 |
US11334623B2 (en) | 2019-03-27 | 2022-05-17 | Western Digital Technologies, Inc. | Key value store using change values for data properties |
US11080239B2 (en) * | 2019-03-27 | 2021-08-03 | Western Digital Technologies, Inc. | Key value store using generation markers |
US11507277B2 (en) | 2019-06-25 | 2022-11-22 | Western Digital Technologies, Inc. | Key value store using progress verification |
CN111399761B (zh) * | 2019-11-19 | 2023-06-30 | 杭州海康威视系统技术有限公司 | 存储资源分配方法、装置及设备、存储介质 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004252663A (ja) * | 2003-02-19 | 2004-09-09 | Toshiba Corp | ストレージ装置、分担範囲決定方法及びプログラム |
JP2010211421A (ja) * | 2009-03-09 | 2010-09-24 | Canon Inc | 管理装置、システム、制御方法、プログラム及び記録媒体 |
WO2012023384A1 (ja) * | 2010-08-19 | 2012-02-23 | 日本電気株式会社 | オブジェクト配置装置及び方法、コンピュータプログラム |
JP2012525634A (ja) * | 2009-04-30 | 2012-10-22 | ネットアップ,インコーポレイテッド | ストライプ化ファイルシステムにおける能力平準化によるデータ分散 |
Family Cites Families (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7036039B2 (en) * | 2002-03-29 | 2006-04-25 | Panasas, Inc. | Distributing manager failure-induced workload through the use of a manager-naming scheme |
JP2006301820A (ja) * | 2005-04-19 | 2006-11-02 | Hitachi Ltd | ストレージシステム及びストレージシステムのデータ移行方法 |
US8190588B1 (en) * | 2005-09-19 | 2012-05-29 | Amazon Technologies, Inc. | Providing a distributed transaction information storage service |
US20110010518A1 (en) * | 2005-12-19 | 2011-01-13 | Srinivas Kavuri | Systems and Methods for Migrating Components in a Hierarchical Storage Network |
US7743276B2 (en) * | 2006-09-27 | 2010-06-22 | Hewlett-Packard Development Company, L.P. | Sufficient free space for redundancy recovery within a distributed data-storage system |
WO2008126202A1 (ja) * | 2007-03-23 | 2008-10-23 | Fujitsu Limited | ストレージシステムの負荷分散プログラム、ストレージシステムの負荷分散方法、及びストレージ管理装置 |
JP4606455B2 (ja) | 2007-12-20 | 2011-01-05 | 富士通株式会社 | ストレージ管理装置、ストレージ管理プログラムおよびストレージシステム |
US8082330B1 (en) * | 2007-12-28 | 2011-12-20 | Emc Corporation | Application aware automated storage pool provisioning |
US7849180B2 (en) * | 2008-04-29 | 2010-12-07 | Network Appliance, Inc. | Load balanced storage provisioning |
US8214404B2 (en) * | 2008-07-11 | 2012-07-03 | Avere Systems, Inc. | Media aware distributed data layout |
JP2010044660A (ja) * | 2008-08-15 | 2010-02-25 | Hitachi Ltd | ストレージシステム及びそのデータ保護方法 |
US8224782B2 (en) * | 2008-09-29 | 2012-07-17 | Hitachi, Ltd. | System and method for chunk based tiered storage volume migration |
JP5526540B2 (ja) | 2008-12-25 | 2014-06-18 | 株式会社リコー | 画像処理装置、アクセス制御方法、アクセス制御プログラム |
JP2011059970A (ja) * | 2009-09-10 | 2011-03-24 | Hitachi Ltd | 外部接続構成におけるボリューム割り当て方法 |
WO2011092738A1 (ja) * | 2010-01-28 | 2011-08-04 | 株式会社日立製作所 | 性能の異なる実領域群で構成されたプールを有するストレージシステムの管理システム及び方法 |
WO2012168967A1 (en) * | 2011-06-07 | 2012-12-13 | Hitachi, Ltd. | Storage apparatus and data management method |
WO2012172601A1 (en) * | 2011-06-14 | 2012-12-20 | Hitachi, Ltd. | Storage system comprising multiple storage control apparatus |
JP2013045379A (ja) | 2011-08-26 | 2013-03-04 | Fujitsu Ltd | ストレージ制御方法、情報処理装置およびプログラム |
US8751657B2 (en) * | 2011-10-04 | 2014-06-10 | Hitachi, Ltd. | Multi-client storage system and storage system management method |
US8566543B2 (en) * | 2011-12-19 | 2013-10-22 | Hitachi, Ltd. | Computer system and reclamation control method |
US8949483B1 (en) * | 2012-12-28 | 2015-02-03 | Emc Corporation | Techniques using I/O classifications in connection with determining data movements |
US10552342B1 (en) * | 2013-03-15 | 2020-02-04 | EMC IP Holding Company LLC | Application level coordination for automated multi-tiering system in a federated environment |
-
2014
- 2014-05-15 JP JP2015518069A patent/JP6222227B2/ja active Active
- 2014-05-15 US US14/781,088 patent/US10708355B2/en active Active
- 2014-05-15 WO PCT/JP2014/002562 patent/WO2014188682A1/ja active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004252663A (ja) * | 2003-02-19 | 2004-09-09 | Toshiba Corp | ストレージ装置、分担範囲決定方法及びプログラム |
JP2010211421A (ja) * | 2009-03-09 | 2010-09-24 | Canon Inc | 管理装置、システム、制御方法、プログラム及び記録媒体 |
JP2012525634A (ja) * | 2009-04-30 | 2012-10-22 | ネットアップ,インコーポレイテッド | ストライプ化ファイルシステムにおける能力平準化によるデータ分散 |
WO2012023384A1 (ja) * | 2010-08-19 | 2012-02-23 | 日本電気株式会社 | オブジェクト配置装置及び方法、コンピュータプログラム |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10860541B2 (en) | 2016-04-11 | 2020-12-08 | Johnson Controls Fire Protection LP | Fire detection system with distributed file system |
Also Published As
Publication number | Publication date |
---|---|
JP6222227B2 (ja) | 2017-11-01 |
US10708355B2 (en) | 2020-07-07 |
JPWO2014188682A1 (ja) | 2017-02-23 |
US20160308965A1 (en) | 2016-10-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6222227B2 (ja) | ストレージノード、ストレージノード管理装置、ストレージノード論理容量設定方法、プログラム、記録媒体および分散データストレージシステム | |
JP6734807B2 (ja) | テナントアウェアストレージシェアリングプラットフォームのための処理装置及びその方法 | |
JP6479639B2 (ja) | 情報処理装置、プログラム、及び、情報処理システム | |
TWI674502B (zh) | 記憶體系統及控制方法 | |
TWI791140B (zh) | 記憶體系統 | |
JP7437117B2 (ja) | ソリッドステートドライブ(ssd)及び分散データストレージシステム並びにその方法 | |
JP6496626B2 (ja) | 異種統合メモリ部及びその拡張統合メモリスペース管理方法 | |
JP4068473B2 (ja) | ストレージ装置、分担範囲決定方法及びプログラム | |
US8566555B2 (en) | Data insertion system, data control device, storage device, data insertion method, data control method, data storing method | |
US11095715B2 (en) | Assigning storage responsibility in a distributed data storage system with replication | |
US11262916B2 (en) | Distributed storage system, data processing method, and storage node | |
JP2020123041A (ja) | メモリシステムおよび制御方法 | |
JP6511795B2 (ja) | ストレージ管理装置、ストレージ管理方法、ストレージ管理プログラムおよびストレージシステム | |
JP7467593B2 (ja) | リソース割振り方法、記憶デバイス、および記憶システム | |
TW201531862A (zh) | 記憶體資料分版技術 | |
US20100161585A1 (en) | Asymmetric cluster filesystem | |
US10387043B2 (en) | Writing target file including determination of whether to apply duplication elimination | |
JP2015022327A (ja) | データ再配置装置、方法およびプログラム | |
US11188258B2 (en) | Distributed storage system | |
JP2013125437A (ja) | 制御装置、プログラムおよびストレージ装置 | |
KR20120063946A (ko) | 대용량 통합 메모리를 위한 메모리 장치 및 이의 메타데이터 관리 방법 | |
US11797338B2 (en) | Information processing device for reading object from primary device specified by identification, information processing system for reading object from primary device specified by identification, and access control method for reading object from primary device specified by identification | |
JP6022116B1 (ja) | 階層化ストレージシステム、ストレージコントローラ及びレプリケーション初期化方法 | |
JP5278254B2 (ja) | ストレージシステム、データ記憶方法及びプログラム | |
Ruty et al. | Collapsing the layers: 6Stor, a scalable and IPv6-centric distributed storage system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14800744 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2015518069 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14781088 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14800744 Country of ref document: EP Kind code of ref document: A1 |