JP7316322B2 - Management system, data rebalance management method, and data rebalance management program - Google Patents

Description

The present invention relates to data rebalancing technology in a storage system configured with a plurality of nodes.

A HCI (Hyper-Converged Infrastructure) system is known in which storage devices mounted on a plurality of servers are integrated by software functions to configure virtual storage on the servers.

In the HCI system, when the I/O load of each node becomes uneven, or when nodes or storage devices are added or replaced, data rebalance processing is executed.

According to the rebalancing process, data in a storage device with a high usage rate is moved to a storage device with a low usage rate.

As a related technique, for example, Patent Document 1 shows a plurality of physical capacities corresponding to a plurality of storage devices including at least one storage device connected to one or more computers included in a computer system and having a compression function. A technique is disclosed for determining instruction contents including a definition related to logical capacity allocation to an entity having a rebalancing function based on capacity information containing information.

WO2018/109816

For example, during the rebalancing process, the I/O performance of each node is degraded, affecting the performance of the system running on each node. In order to avoid the impact of such rebalancing processing on the performance of the operating system, users often manually execute rebalancing processing when the system is stopped, such as at night or on holidays, increasing the operational load on the user. there is

The present invention has been made in view of the circumstances described above, and an object of the present invention is to provide a technique capable of appropriately rebalancing data.

In order to achieve the above object, a management system according to one aspect has a plurality of nodes that control data input/output to and from storage devices to be controlled, and manages predetermined data units by redundantly storing them in the plurality of storage devices. A management system for managing data rebalancing in a storage system, said management system having a processor unit, said processor unit acquiring data input/output load information of a plurality of said nodes, as a transmission source of the data unit to be rebalanced, based on the load information, one node is determined from among the plurality of nodes, and the data unit to be rebalanced is transmitted from the determined node to a predetermined Send to the destination node.

According to the present invention, data rebalancing can be appropriately performed.

FIG. 1 is an overall configuration diagram of a computer system according to one embodiment. FIG. 2 is a configuration diagram of a computer according to one embodiment. FIG. 3 is a configuration diagram of a backup system according to one embodiment. FIG. 4 is a configuration diagram of a management system according to one embodiment. FIG. 5 is a configuration diagram of a rebalance priority table according to one embodiment. FIG. 6 is a configuration diagram of a rebalance route table according to one embodiment. FIG. 7 is a configuration diagram of a depletion condition table according to one embodiment. FIG. 8 is a configuration diagram of a node capacity table according to one embodiment. FIG. 9 is a configuration diagram of a node I/O table according to one embodiment. FIG. 10 is a configuration diagram of a backup information table according to one embodiment. FIG. 11 is a flowchart of table update processing according to one embodiment. FIG. 12 is a flowchart of capacity and I/O information notification processing according to one embodiment. FIG. 13 is a flowchart of backup information notification processing according to one embodiment. FIG. 14 is a flowchart of rebalance control processing according to one embodiment.

Embodiments will be described with reference to the drawings. It should be noted that the embodiments described below do not limit the invention according to the scope of claims, and that all of the elements described in the embodiments and their combinations are essential to the solution of the invention. is not limited.

Further, in the following description, reference numerals may be used when similar elements are not distinguished, and element IDs may be used when similar elements are distinguished. For example, when the nodes are not distinguished, they are referred to as "node 220", and when the nodes are distinguished, they are referred to as "node n1", "node n2", and "node n3".

Also, in the following description, "interface section" includes one or more interfaces. The one or more interfaces may be one or more similar interface devices (e.g., one or more NICs (Network Interface Cards)) or two or more different types of interface devices (e.g., NIC and HBA (Host Bus Adapter)). There may be.

Also, in the following description, a “storage unit” includes one or more memories. The at least one memory may be volatile memory or non-volatile memory. The storage unit is mainly used during processing by the processor unit.

Also, in the following description, a "processor unit" includes one or more processors. The at least one processor is typically a microprocessor such as a CPU (Central Processing Unit). Each of the one or more processors may be single-core or multi-core. A processor may include hardware circuitry that performs some or all of the processing.

Also, in the following description, the information may be described using an expression such as "AAA table", but the information may be expressed in any data structure. That is, the "AAA table" can be referred to as "AAA information" to indicate that the information does not depend on the data structure. Also, in the following description, the configuration of each table is an example, and one table may be divided into two or more tables, or all or part of two or more tables may be one table. good.

In the following description, the processing unit (function) may be described using the expression “kkk unit”, but the processing unit may be realized by executing one or more computer programs by the processor unit. Alternatively, it may be implemented by one or more hardware circuits (eg, FPGA or ASIC (Application Specific Integrated Circuit)). When the program is realized by the processor unit, the predetermined processing is performed while appropriately using storage resources (e.g., memory) and/or communication interface devices (e.g., communication ports). It may be at least part of the part. The processing described with the processing unit as the main body of operation may be processing performed by the processor unit or a device having the processor unit. Also, the processor unit may include hardware circuitry that performs part or all of the processing. Programs may be installed on the processor from program sources. The program source may be, for example, a program distribution computer or a computer-readable recording medium (for example, a non-temporary recording medium). The description of each processing unit is an example, and a plurality of processing units may be combined into one processing unit, or one processing unit may be divided into a plurality of processing units.

Further, in the following explanation, the processing may be explained with the "program" as the main body of operation. , the operator of the processing may be a processor (or a computer having a processor). The program may be installed on the computer from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium. Also, in the following description, two or more programs may be implemented as one program, and one program may be implemented as two or more programs.

FIG. 1 is an overall configuration diagram of a computer system according to one embodiment.

The computer system 1 includes a storage system 20, a management system 100, a backup system 300, and a backup storage device 400 as an example of a backup device. The storage system 20, the management system 100, the backup system 300, and the backup storage device 400 are connected via the network 10 so as to be able to communicate with each other. The network 10 is, for example, an IP (Internet Protocol) network. Note that the backup storage device 400 may be configured on the backup system 300, and in short, any device that can access backup data via the network 10 may be used.

Management system 100 runs manager 110 .

Backup system 300 runs a backup manager 310 . The backup storage device 400 stores volume data (backup data) 410 of the storage system 20 backed up by the backup system 300 .

The storage system 20 includes one or more computers 200 (for example, computers A and B) and includes multiple nodes 220 . Computer 200 includes one or more nodes 220 and one or more storage devices 210 . Node 220 may be an example of a computer program executed on computer 200 . The number of nodes 220 of the computer 200 and the number of storage devices 210 are the same, and correspond one-to-one. In FIG. 1, nodes n1-n3 correspond to storage devices A-C, respectively. In the storage system 20, the same volume (an example of data unit) is stored in multiple storage devices 210 corresponding to multiple nodes 220 in order to ensure redundancy.

FIG. 2 is a configuration diagram of a computer according to one embodiment.

The computer 200 has a storage device 210 , a processor 211 , a memory 212 and a network I/F (network interface) 213 .

The storage device 210 is typically a physical non-volatile storage device such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). The storage device 210 stores various data.

The network I/F 213 is, for example, an interface such as a wired LAN card or a wireless LAN card, and communicates with other devices (eg, management system 100, backup system 300) via the network 10. FIG.

Processor 211 executes various processes according to programs stored in memory 212 and/or storage device 210 .

The memory 212 is an example of a storage unit and stores the node 220 . Node 220 is executed by processor 211 .

The node 220 provides the logical storage space of the storage device 210 corresponding to the node 220 to applications within the computer 200 and applications outside the computer 200, for example. The node 220 accepts an I/O request to the logical storage space, and according to the I/O request, executes data write or read to the storage device 210 corresponding to that node 220 (an I/O command is sent to the storage device). 210). A node 220 sends write-targeted data (I/O request for write-targeted data) to its corresponding storage device 210, for example, to maintain data redundancy. For example, it can be transferred to a node 220 in another computer 200). The node 220 may be SDS (Software Defined Storage), for example. Node 220 has a storage service 221 .

The storage service 221 corresponds to a virtual storage controller which is a virtual controller that accepts I/O requests and writes or reads data from the storage device 210 according to the I/O requests. The storage service 221 includes a capacity monitoring unit 222, a capacity notification unit 223, an I/O monitoring unit 224, an I/O notification unit 225, a rebalancing unit 226, a data difference management unit 227, and a backup information monitoring unit. 228 and a backup information notification unit 229 .

The capacity monitoring unit 222 monitors (for example, periodically checks) the logical capacity and physical capacity of the storage device 210 corresponding to the node 220 to which it belongs (referred to as its own node). The capacity notification unit 223 notifies the manager 110 of the logical capacity and physical capacity specified by the capacity monitoring unit 222 .

The I/O monitoring unit 224 monitors the I/O load on the storage device 210 of its own node. The I/O notification unit 225 notifies the manager 110 of the I/O load specified by the I/O monitoring unit 224 . The rebalancing unit 226 executes rebalancing processing for transferring data according to instructions from the manager 110 . The data difference management unit 227 manages information (difference cross section information) indicating a difference cross section from data (volume) at a predetermined point in time. The backup information monitoring unit 228 monitors backup of data in the storage device 210 of its own node. The backup information notification unit 229 notifies the manager 110 of the backup information specified by the backup information monitoring unit 228 .

FIG. 3 is a configuration diagram of a backup system according to one embodiment.

Backup system 300 has processor 301 , memory 302 , input device 303 , output device 304 and network I/F 305 .

The input device 303 is, for example, a keyboard and pointing device, and receives various inputs. The output device 304 is, for example, a display device such as a liquid crystal display, and displays and outputs various information. The input device 303 and the output device 304 may be integrated as a touch panel, for example.

The network I/F 305 is, for example, an interface such as a wired LAN card or wireless LAN card, and communicates with other devices (eg, the computer 200, the backup storage device 400) via the network 10. FIG.

The processor 301 executes various processes according to programs stored in the memory 302 .

The memory 302 is an example of a storage unit and stores a backup manager 310 . Backup manager 310 is executed by processor 301 .

Backup manager 310 has a backup service 320 .

The backup service 320 performs processing for storing backup data of data stored in the storage device 210 of the storage system 20 in the backup storage device 400 . Data backup is performed, for example, in units of volumes (an example of data units). The backup service 320 has a backup information notifier 321 . The backup information notification unit 321 transmits information (backup information) regarding backup of data stored in the storage device 210 to the computer 200 . The backup information includes an identifier (volume ID) of the backed up volume, data differential section information of the backed up data at the time of backup, and a backup hash.

FIG. 4 is a configuration diagram of a management system according to one embodiment.

Management system 100 has processor 101 , memory 102 , input device 103 , output device 104 and network I/F 105 .

The input device 103 is, for example, a keyboard and pointing device, and receives various inputs. The output device 104 is, for example, a display device such as a liquid crystal display, and displays and outputs various information. The input device 103 and the output device 104 may be integrated as a touch panel, for example.

The processor 101 executes various processes according to programs stored in the memory 102 .

Memory 102 is an example of a storage unit and stores manager 110 . Manager 110 is executed by processor 101 .

Manager 110 has a data management service 111 .

The data management service 111 acquires various information from the node 220, determines the transmission route (at least the source node) of the data unit to be rebalanced based on the various information, and rebalances the data corresponding to the determined content. A process of transmitting a process instruction to the node 220 is executed. The data management service 111 includes a table setting unit 118, a capacity information reception unit 119, an I/O information reception unit 120, a depletion detection unit 121, a backup information reception unit 122, a backup data detection unit 123, a rebalance It has a selection unit 124 and a rebalance instruction unit 125 . The data management service 111 also manages a rebalance priority table 112, a rebalance path table 113, an exhaustion condition table 114, a node capacity table 115, a node I/O table 116, and a backup information table 117. do.

A table setting unit 118 sets and updates the tables 112 to 114 . The capacity information receiving unit 119 receives capacity information of the storage device 210 from each node 220 and registers it in the node capacity table 115 . The I/O information receiving unit 120 receives the I/O information of the storage device 210 from each node 220 and registers it in the node I/O table 116 . The exhaustion detection unit 121 detects capacity exhaustion of the storage device 210 of each node 220 . The backup information receiving unit 122 receives backup information of data backed up from the backup system 300 and registers it in the backup information table 117 . The backup data detection unit 123 detects whether backup data of data exists. The rebalance selection unit 124 performs processing for selecting a path for transferring data of a rebalance target volume. The rebalance instruction unit 125 transmits an instruction to the rebalance unit 226 of the node 220 to transfer the volume through the route selected by the rebalance selection unit 124 .

FIG. 5 is a configuration diagram of a rebalance priority table according to one embodiment.

The rebalance priority table 112 is a table for managing the priority of rebalance processing, and stores an entry (row) for each rebalance processing. An entry in rebalance priority table 112 has columns for rebalance ID 112a, volume ID 112b, and priority 112c.

An identifier (rebalance ID) for identifying a rebalance process is stored in the rebalance ID 112a. The volume ID 112b stores the identifier (volume ID) of the volume to be rebalanced corresponding to the entry. The priority 112c stores the priority of rebalancing processing corresponding to the entry. In this embodiment, the smaller the priority value, the higher the priority.

For example, according to the first line, the rebalancing process identified by the rebalancing ID r1 targets the volumes with volume IDs v2 and v3, and the priority of the rebalancing process is "2". is shown. Also, according to the first and second rows, the rebalancing process for the volumes with volume IDs v2 and v3 has a higher priority than the rebalancing process for the volume with volume ID v1 (value is small), the rebalancing process for the volume with the volume ID of v1 is preferentially performed.

FIG. 6 is a configuration diagram of a rebalance route table according to one embodiment.

The rebalance route table 113 is a table for managing routes to be rebalanced, and stores an entry for each route. A path includes at least nodes, and may include cables, physical ports of communication I/F, and virtual ports. An entry in the rebalance path table 113 has columns of node ID 113a, volume ID 113b, and path ID 113c.

The node ID 113a stores the identifier (node ID) of the node 220 belonging to the path corresponding to the entry. Note that in this embodiment, an ID indicating the backup storage device 400 is also stored in the node ID 113a, although it is not the node 220. FIG. In FIG. 6, the node ID including b, such as b1, indicates the ID of the backup storage device. The volume ID 113b stores an ID (volume ID) of a volume that can be rebalanced (transferred) on the route corresponding to the entry. The path ID 113c stores an ID (path ID) of the path corresponding to the entry. Here, the path includes the node 220 and may include a cable, a physical I/F port, or a virtual port.

In the rebalance path table 113, multiple paths are registered for the same volume.

For example, the second line indicates that the volume with the volume ID of v3 has a path with the path ID of e5 that connects to the node 220 (node n1) with the node ID of n1. , there is a path with a path ID of e10 that connects the volume with a volume ID of v3 to the node 220 (node n3) with a node ID of n3. According to these lines, it can be seen that the volume with volume ID v3 can be transferred through either the e5 route leading to node n1 or the e10 route leading to node n3.

FIG. 7 is a configuration diagram of a depletion condition table according to one embodiment.

The exhaustion condition table 114 stores information on exhaustion conditions. The depletion condition table 114 stores entries for each node 220 . The entry in the exhaustion condition table 114 has columns of node ID 114a, capacity exhaustion condition 114b, and I/O bandwidth exhaustion condition 114c.

The node ID 114a stores the node ID of the node 220 corresponding to the entry. The capacity exhaustion condition 114b stores an exhaustion condition (capacity exhaustion condition) under which it is determined that the storage capacity of the storage device 210 in the node 220 corresponding to the entry is exhausted. The I/O band depletion condition 114c stores a depletion condition (I/O band depletion condition) under which it is determined that the I/O band in the node 220 corresponding to the entry is depleted.

For example, according to the first row, the capacity exhaustion condition of node n1 is that the remaining physical capacity (physical remaining capacity) of the storage device 210 is less than 10 GB, and the I/O band exhaustion condition is I/O bandwidth of node n1. This indicates that the remaining O band (remaining I/O band amount) is less than 10 kbps.

FIG. 8 is a configuration diagram of a node capacity table according to one embodiment.

The node capacity table 115 stores information on the logical capacity and physical capacity of the storage device 210 corresponding to the node 220 . The node capacity table 115 stores entries for each node. The entry of the node capacity table 115 has columns of node ID 115a, logical used capacity 115b, logical remaining capacity 115c, physical used capacity 115d, and physical remaining capacity 115e.

The node ID 115a stores the node ID of the node 220 corresponding to the entry. The logical used capacity 115b stores the used capacity (logical used capacity) in the logical capacity (logical capacity) of the storage area of the storage device 210 corresponding to the node 220 corresponding to the entry. The logical remaining capacity 115c stores the remaining unused capacity (logical remaining capacity) in the logical capacity of the storage device 210 corresponding to the node 220 corresponding to the entry. The physical used capacity 115d stores the used capacity (physical used capacity) in the physical capacity (physical capacity) of the storage area of the storage device 210 corresponding to the node 220 corresponding to the entry. The physical remaining capacity 115e stores the unused remaining capacity (physical remaining capacity) in the physical capacity of the storage device 210 corresponding to the node 220 corresponding to the entry.

For example, the first row indicates that node n1 has a logical used capacity of 400 GB, a logical remaining capacity of 200 GB, a physical used capacity of 60 GB, and a physical remaining capacity of 140 GB.

FIG. 9 is a configuration diagram of a node I/O table according to one embodiment.

The node I/O table 116 stores I/O load information for each path. The node I/O table 116 stores an entry for each path. An entry in the node I/O table 116 has columns of path ID 116a, node ID 116b, I/O bandwidth usage 116c, and remaining I/O bandwidth 116d.

The path ID of the path corresponding to the entry is stored in the path ID 116a. The node ID 116b stores the node ID of the node 220 included in the path corresponding to the entry. The I/O band usage 116c stores the I/O band usage (I/O band usage) in the path corresponding to the entry. The remaining I/O bandwidth amount 116d stores the remaining I/O bandwidth amount (remaining I/O bandwidth amount) in the path corresponding to the entry.

For example, according to the first row, path e1 includes node n1, the I/O bandwidth usage is 300 kbps, and the remaining I/O bandwidth is 0 kbps.

FIG. 10 is a configuration diagram of a backup information table according to one embodiment.

The backup information table 117 stores volume information (backup information) backed up by the backup system 300 . The backup information table 117 stores an entry for each volume backup process. The entry of the backup information table 117 has columns of a backup ID 117a, a volume ID 117b, a data difference section information 117c, and a backup hash 117d.

The backup ID 117a stores an identifier (backup ID) corresponding to the backup process corresponding to the entry. The volume ID 117b stores the volume ID of the volume to be backed up corresponding to the entry. The data difference cross-section information 117c stores information indicating information (data difference cross-section information) capable of specifying a difference portion from the state of the volume corresponding to the entry at a predetermined point in time. By comparing this data difference cross-sectional information, it is possible to grasp the portion where there is a difference with respect to the same volume. The backup hash 117d stores a hash value for the volume corresponding to the entry.

For example, according to the first line, in the backup process with the backup ID of b1, the volume with the volume ID of v2 is backed up, the data differential section information of that volume is data of 20200105_v0002_1092131, and the hash value of the volume is 78tffa978s3j. indicates that there is

Next, processing operations in the computer system 1 will be described.

First, table update processing by the management system 100 will be described.

FIG. 11 is a flowchart of table update processing according to one embodiment. Table update processing is performed when settings are updated by a user (eg, administrator).

The table setting unit 118 receives input of information (for example, node ID) about each node 220 from the user via the input device 103 (step S11).

The table setting unit 118 receives input of information regarding the depletion condition of the storage device 210 corresponding to the node 220 from the user via the input device 103 (step S12).

The table setting unit 118 registers the information received in steps S11 and S12 in the depletion condition table 114, that is, updates the depletion condition table 114 (step S13). At this time, since the storage device 210 and the node 220 correspond one-to-one, the table setting unit 118 can uniquely determine the node 220 from the information regarding the storage device 210 . Initial values may be registered in advance in the capacity depletion condition 114b and the I/O band depletion condition 114c. In this case, step S12 may be skipped.

The table setting unit 118 receives, from the user via the input device 103, information (for example, rebalance ID, volume ID) regarding the functions of the rebalance unit 226 of the node 220 (step S14).

The table setting unit 118 receives, from the user via the input device 103, information (for example, a priority value) regarding the priority of rebalancing processing by the rebalancing unit 226 (step S15).

The table setting unit 118 registers the information received in steps S14 and S15 in the rebalance priority table 112, that is, updates the rebalance priority table 112 (step S16). Note that the rebalance ID and volume ID may be registered in the rebalance priority table 112 in advance. In this case, step S14 may be skipped.

The table setting unit 118 receives, from the user via the input device 103, information (for example, the volume ID and the ID of the backup storage device 400 to which the volume is backed up) regarding the backup data of the volume of the node 220 (step S17).

The table setting unit 118 receives, from the user via the input device 103, information (for example, a path ID) regarding a path (rebalancing path) in volume rebalancing processing (step S18).

The table setting unit 118 registers the information received in steps S17 and S18 in the rebalance route table 113, that is, updates the rebalance route table 113 (step S19).

Next, the capacity and I/O information notification processing performed by the storage service 221 of each node 220 of the computer 200 will be described.

FIG. 12 is a flowchart of capacity and I/O information notification processing according to one embodiment. This capacity and I/O information notification process is performed periodically, for example. 12, the node 220 including the storage service 221 is called the target node 220, and the storage device 210 corresponding to the target node 220 is called the target storage device 210. FIG.

The capacity monitoring unit 222 requests capacity information from the target storage device 210 (step S21), and receives the logical used capacity, logical remaining capacity, physical used capacity, and physical remaining capacity of the target storage device 210 as capacity information from the target storage device 210. Information indicating capacity is received (step S22).

The I/O monitoring unit 224 acquires, from the target node 220, the I/O bandwidth usage amount and the remaining I/O bandwidth amount in each path of the target node 220 as the I/O information on the target node 220 (step S23). .

The capacity notification unit 223 and the I/O notification unit 225 transmit the capacity information received by the capacity monitoring unit 222 in step S22 and the I/O information acquired by the I/O monitoring unit 224 in step S23 to the manager of the management system 100. 110 (step S24). It should be noted that the manager 110 of the management system 100 performs the following update processing every time it receives capacity information and I/O information. That is, the capacity information receiving unit 119 registers the received capacity information (information indicating logical used capacity, logical remaining capacity, physical used capacity, and physical remaining capacity) in the node capacity table 115, and the I/O information receiving unit 120 registers the received I/O information (I/O bandwidth usage and remaining I/O bandwidth in each path of the target node 220) in the node I/O table 116. FIG. As a result, the node capacity table 115 and the node I/O table 116 in the management system 100 reflect the latest state of each node 220 .

Next, backup information notification processing performed by the storage service 221 of each node 220 of the computer 200 will be described.

FIG. 13 is a flowchart of backup information notification processing according to one embodiment. This backup information notification process is performed periodically, for example.

The backup information monitoring unit 228 requests the backup information from the backup system 300 (step S41), and receives the backup information from the backup manager 310 of the backup system 300 as the backup ID corresponding to the backed up volume, the data differential section information, and the backup information. A hash is received (step S42).

The backup information notification unit 229 notifies the manager 110 of the management system 100 of the backup information received by the backup information monitoring unit 228 in step S42 (step S43). It should be noted that the manager 110 of the management system 100 performs the following updating process each time it receives backup information. That is, the backup information receiving unit 122 registers the received backup information in the backup information table 117. FIG. As a result, in the management system 100 , the backup information table 117 reflects the latest state of the volumes backed up by the backup system 300 .

Next, rebalancing control processing performed by the manager 110 of the management system 100 will be described.

FIG. 14 is a flowchart of rebalance control processing according to one embodiment. The rebalance control process may be performed periodically. Further, the rebalance control process may be executed when the difference in the physical capacity usage rate (physical capacity usage rate) of the storage device 2210 corresponding to each node 220 exceeds a predetermined threshold. For example, if the threshold is 20% and the physical capacity usage rates of node n1, node n2, and node n3 are 20%, 30%, and 50%, the difference in physical capacity usage rate between node n1 and node n3 is 30%. Yes, it exceeds the threshold, so rebalance control processing may be executed.

The depletion detection unit 121 determines whether or not there is an unexecuted node, which is a node that has not performed the depletion detection process (step S51). As a result, if there is no incomplete node (step S51: NO), the depletion detection unit 121 terminates the rebalance control process.

On the other hand, if there is an unimplemented node (step S51: YES), the depletion detection unit 121 checks whether or not the capacity of the storage device 210 corresponding to the unimplemented node is exhausted. Refer to it and judge (step S52).

For example, in the depletion condition table 114 of FIG. 7, from the depletion condition of node n1, node n1 is determined to be depleted when the remaining physical capacity is less than 10 GB. In the node capacity table 115, node n1 has a logical remaining capacity of 200 GB and a physical remaining capacity of 140 GB. Since the exhaustion condition for node n1 is not satisfied, node n1 is not determined to be exhausted. On the other hand, for example, in the exhaustion condition table 114, from the exhaustion condition of the node n2, the node n2 is determined to be exhausted when the remaining physical capacity is less than 10 GB. In the node capacity table 115, the logical remaining capacity of node n2 is 200 GB, and the physical remaining capacity is 8 GB. Since the exhaustion condition for node n2 is satisfied, node n2 is determined to be exhausted.

If depletion is not detected in step S52 (step S52: NO), the depletion detection unit 121 advances the process to step S51.

On the other hand, if depletion is detected in step S52 (step S52: YES), the rebalance selection unit 124 refers to the depletion condition table 114, the rebalance route table 113, and the rebalance priority table 112, and A target rebalancing process is searched, and it is determined whether or not the rebalancing process can be performed (step S53). In this way, when depletion is detected, search for rebalancing processing proceeds, so it is possible to avoid running rebalancing processing even when rebalancing is unnecessary, such as when depletion is not detected.

For example, when the depletion of node n2 is detected in step S52, the rebalance selection unit 124 finds from the rebalance path table 113 that volume v1 is stored in the storage device 210 corresponding to node n2. Also, from the rebalance priority table 112, it can be seen that volume v1 can be rebalanced by rebalancing process r2.

If it is determined in step S53 that the rebalancing process cannot be performed (step S53: NO), the rebalancing selection unit 124 notifies the user of a warning that the capacity is insufficient (for example, displays the warning on the output device 104) ( Step S61), the process proceeds to step S51.

On the other hand, if an executable rebalance process is found (step S53: YES), the rebalance selection unit 124 refers to the rebalance priority table 112 and selects the rebalance process to be performed (step S54). For example, if it is possible to perform rebalance processing on volume v1 and volume v3, referring to the priority 112c of the rebalance priority table 112, the rebalance processing for volume v1 (rebalance ID r2) is performed. is "1", the priority of the rebalance process (rebalance IDr1) for volume v3 is "2", and the priority of the rebalance process (rebalance IDr2) is the highest. . Therefore, the rebalancing selection unit 124 selects the rebalancing process with the rebalancing ID r2. In this way, when a plurality of rebalancing processes can be performed, the rebalancing process with the highest priority can be preferentially selected.

Next, the rebalancing selection unit 124 refers to the rebalancing path table 113 and the node I/O table 116, and according to the setting information for the rebalancing process, rebalances the target volume of the rebalancing process (target volume). to select the route to be used (step S55).

Here, the setting information for the rebalancing process is, for example, information set in advance by the user. There is a speed priority setting (second setting) that prioritizes the speed (transfer processing efficiency) until the node 220 becomes available.

In the case of performance priority setting, the rebalance selection unit 124 refers to the rebalance path table 113 to identify paths including the target volume, refers to the node I/O table 116, and selects an I in the identified paths. Select one path with the lowest /O load (for example, the highest amount of remaining I/O bandwidth). As a result, the node 220 from which the target volume is transferred is determined.

On the other hand, in the case of speed priority setting, the rebalance selection unit 124 refers to the rebalance path table 113, identifies all paths including the target volume, and if there are multiple paths, Choose a route. Here, the plurality of routes to be selected may be all of the specified routes or may be a part of the plurality of routes. Also, the rebalance selection unit 124 may select a plurality of paths whose I/O load is equal to or less than a predetermined load from among the specified paths. As a result, multiple nodes 220 from which the target volume is transferred are determined.

Next, the rebalancing selection unit 124 determines the instruction content for executing the rebalancing process selected in step S54 through the route determined in step S55 (step S56). When the same volume is rebalanced by a plurality of paths, the contents of the instruction to transfer different ranges of the same volume in parallel are determined for each path. In addition, in the contents of the instruction, the destination node of the target volume may be a predetermined node, for example, a node with a large physical remaining capacity.

Next, the rebalance selection unit 124 determines whether or not the content of the instruction includes the content of using the backup data (step S57).

As a result, if the content of the instruction does not include the content of using the backup data (step S57: NO), the rebalancing selection unit 124 advances the process to step S59, while the content of the instruction indicates that the backup data is used. If the content is included (step S57: YES), the rebalance selection unit 124 gives an instruction (difference catch-up Instruction: For example, including data difference sectional information of backup data (step S58), the process proceeds to step S59.

In step S59, the rebalance instruction unit 125 transmits the instruction content created by the rebalance selection unit 124 to the node 220 involved in the rebalancing process (for example, the node serving as the data transmission source on the path). .

The rebalancing unit 226 of the node 220 that has received the content of the instruction executes the rebalancing process according to the content of the instruction. For example, the rebalancing unit 226 transfers the target volume to the transfer destination through a route according to the content of the instruction. Also, if the instruction content includes a difference catch-up instruction, the data difference management unit 227 stores the data difference section information included in the instruction and the current data difference section information of the target volume corresponding to the backup data. Based on this, the difference between the target volume and the backup data is identified, the data corresponding to the difference is transmitted to the rebalancing destination, and updated with the data corresponding to the difference. The rebalancing unit 226 returns rebalancing completion to the manager 110 when the rebalancing is completed.

When the rebalancing instruction unit 125 of the manager 110 receives the rebalancing end notification from the rebalancing unit 226 (step S60), the process proceeds to step S51.

According to the rebalance control process described above, when performance priority is set, the target volume is transferred using one path with the least I/O load (for example, node 220). As a result, it is possible to reduce the impact of performance degradation due to rebalancing processing in the storage system 20 . Also, if the target volume is transferred through a path that includes the backup storage device 400 instead of the node 220, it is possible to prevent the performance of the storage system 20 from deteriorating.

Also, when speed priority is set, different parts of the target volume are transferred in parallel by a plurality of paths including a plurality of nodes 220 . As a result, the target volume can be quickly transferred to the transfer destination node 220, and the target volume of the transfer destination node 220 can be quickly made available.

Note that the rebalance control process is not limited to being executed periodically. You may make it perform when it exchanges. As the rebalance control process in this case, for example, the processes of steps S54 to S60 may be repeatedly executed for each volume to be transferred to the new node 220 by the rebalance process.

It should be noted that the present invention is not limited to the above-described embodiments, and can be modified appropriately without departing from the scope of the present invention.

For example, in the above embodiment, the computer system 1 is provided with the backup system 300 and the backup storage device 400, but the backup system 300 and the backup storage device 400 may not be provided.

In the above-described embodiment, as rebalancing settings, a performance priority setting that prioritizes the performance of a node and a speed priority setting that prioritizes the speed until a node to which data is transferred can be used are selected. Although it can be set, the present invention is not limited to this, and it may be fixedly set to either the performance priority setting or the speed priority setting.

1 computer system 10 network 20 storage system 100 management system 101 processor 110 manager 200 computer 210 storage device 220 node 300 backup system 400 backup storage device

Claims (8)

A management system that manages data rebalancing in a storage system that has a plurality of nodes that control data input/output to and from controlled storage devices, and that redundantly stores and manages predetermined data units in a plurality of storage devices. hand,
The management system has a processor unit,
The processor unit
obtaining data input/output load information of the plurality of nodes;
determining one node from among the plurality of nodes based on the load information as a transmission source of a predetermined rebalancing target data unit;
causing the data unit to be rebalanced to be transmitted from the determined node to a predetermined transfer destination node ;
The processor unit
Receiving a setting of either a first setting emphasizing maintenance of performance of each storage system or a second setting emphasizing transfer processing efficiency of the data unit for the transfer of data units to be rebalanced,
determining one node from among the plurality of nodes based on the load information as a transmission source of a predetermined data unit to be rebalanced when the first setting is accepted;
a plurality of nodes among a plurality of nodes connected to a storage device storing the data unit to be rebalanced as a transmission source of the predetermined data unit to be rebalanced when the second setting is accepted; determine the node,
When a plurality of nodes are determined as a transmission source of a predetermined data unit to be rebalanced, different portions of the data unit to be rebalanced are transmitted in parallel from the plurality of nodes to a predetermined transfer destination node. let
management system. The processor unit
A plurality of storage devices connected to the storage device storing the data unit to be rebalanced based on the load information as a transmission source of the predetermined data unit to be rebalanced when the second setting is accepted 2. The management system of claim 1 , wherein the plurality of nodes are determined from among the nodes of the . The processor unit
3. The management system according to claim 2 , wherein the load information determines a plurality of nodes having a predetermined load or less as sources of data units to be rebalanced. The load information is load information for each route that can be accessed for each data unit,
The processor unit
Based on the load information, as a route for transmitting a predetermined data unit to be rebalanced, one route is determined from among routes capable of transmitting the data unit to be rebalanced, based on the load information;
2. The management system according to claim 1, wherein the data unit to be rebalanced is read by the determined route and transmitted to a predetermined transfer destination node. Data at a predetermined point in time for data units stored in the storage system is stored in a predetermined backup device,
The processor unit acquires backup information indicating data units stored in the backup device,
determining whether or not the data unit to be rebalanced is stored in the backup device based on the backup information;
when data of a predetermined rebalance target data unit at a predetermined point in time is stored in a backup device, determining the backup device as a transmission source of the predetermined rebalance target data unit;
2. The method according to claim 1, wherein the data is read from the backup device and stored in a storage device connected to the transfer destination node, and the stored data reflects the change difference of the data unit from the predetermined point in time. management system. Data in units of data at a predetermined point in time for data units stored in the storage system is stored in a predetermined backup device,
The processor unit acquires backup information indicating data units stored in the backup device,
When data of a predetermined rebalancing target data unit at a predetermined point in time is stored in a backup device, the load information and the input/output load of the backup device are used as a transmission source of the predetermined rebalancing target data unit. determining one of the plurality of nodes and the backup device based on the information;
When the backup device is determined as the transmission source, it is read from the backup device and stored in a storage device connected to the transfer destination node, and the stored data is read from the predetermined point in time. 2. The management system according to claim 1, wherein a change difference in units of data is reflected. Data from a management system that manages rebalancing of data by a storage system that has multiple nodes that control data input/output to and from storage devices to be controlled, and manages by redundantly storing predetermined data units in multiple storage devices. A rebalancing management method comprising:
The management system is
Acquiring data input/output load information of the plurality of nodes;
determining one node from among the plurality of nodes based on the load information as a transmission source of a predetermined rebalancing target data unit;
causing the data unit to be rebalanced to be transmitted from the determined node to a predetermined transfer destination node ;
Receiving a setting of either a first setting emphasizing maintenance of performance of each storage system or a second setting emphasizing transfer processing efficiency of the data unit for the transfer of data units to be rebalanced,
determining one node from among the plurality of nodes based on the load information as a transmission source of a predetermined data unit to be rebalanced when the first setting is accepted;
a plurality of nodes among a plurality of nodes connected to a storage device storing the data unit to be rebalanced as a transmission source of the predetermined data unit to be rebalanced when the second setting is accepted; determine the node,
When a plurality of nodes are determined as a transmission source of a predetermined data unit to be rebalanced, different portions of the data unit to be rebalanced are transmitted in parallel from the plurality of nodes to a predetermined transfer destination node. let
Data rebalance management method. It is executed by a computer that manages data rebalancing in a storage system that has a plurality of nodes that control data input/output to and from a storage device to be controlled, and that redundantly stores and manages predetermined data units in a plurality of storage devices. A data rebalance management program that
to the computer;
acquire data input/output load information of the plurality of nodes;
determining one node from among the plurality of nodes based on the load information as a transmission source of a predetermined rebalance target data unit;
causing the data unit to be rebalanced to be transmitted from the determined node to a predetermined transfer destination node ;
accepting either a first setting emphasizing performance maintenance of each storage system or a second setting emphasizing transfer processing efficiency of the data unit for the transfer of data units to be rebalanced;
determining one node from among the plurality of nodes based on the load information as a transmission source of a predetermined data unit to be rebalanced when the first setting is accepted;
a plurality of nodes among a plurality of nodes connected to a storage device storing the data unit to be rebalanced as a transmission source of the predetermined data unit to be rebalanced when the second setting is accepted; let the node decide,
When a plurality of nodes are determined as a transmission source of a predetermined data unit to be rebalanced, different portions of the data unit to be rebalanced are transmitted in parallel from the plurality of nodes to a predetermined transfer destination node. let
Data rebalance management program.