Information processing system and configuration management method

The information processing system with redundancy group management and active-standby control units addresses the challenge of maintaining high IO performance and reliability by ensuring straight paths and bindings, even during storage node changes.

JP7590316B2Active Publication Date: 2024-11-26HITACHI VANTARA LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
JP2021209695
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2021-12-23
Publication Date
2024-11-26
Estimated Expiration
2041-12-23

AI Technical Summary

Technical Problem

Existing information processing systems using SCSI architecture models face issues with maintaining high IO performance and reliability when storage configurations change due to the addition or removal of storage nodes, leading to degraded performance and reliability as IO processing is forced through external networks.

Method used

An information processing system with multiple storage nodes connected via a network, where logical units are managed in redundancy groups, and control units are configured in active-standby pairs to ensure data control redundancy and efficient IO processing, maintaining straight paths and bindings even during node failures.

Benefits of technology

The system achieves data control redundancy, suppressing declines in IO performance and reliability by ensuring straight paths and bindings, even during system operation changes, thus maintaining high performance and reliability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007590316000001
    Figure 0007590316000001
  • Figure 0007590316000002
    Figure 0007590316000002
  • Figure 0007590316000003
    Figure 0007590316000003
Patent Text Reader

Abstract

To implement redundancy of data control and suppress deterioration of IO performance and reliability during system operation in an information processing system that provides a virtual volume using a plurality of storage nodes.SOLUTION: In an information processing system 1, a plurality of control units (control software 53) implemented in different storage nodes 3 are managed as a redundancy group, and one or more first logical units (SLUs 52) are associated with the redundancy group. The plurality of control units included in the redundancy group include an active control unit that processes an IO request with respect to the first logical unit, and a standby control unit that takes over the processing when the active control unit fails. A processor (configuration management unit 55) arranges at least one second logical unit (ALU 51) in the storage node 3 in which the active control unit exists, and associates the second logical unit with the first logical unit controlled by the active control unit.SELECTED DRAWING: Figure 4
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an information processing system and a configuration management method, and is suitable for application to an information processing system that provides a virtual volume from the storage of a plurality of nodes and a configuration management method thereof. [Background technology]

[0002] In recent years, attention has been focused on SDS (Software Defined Storage), which is constructed by implementing storage control software in a general-purpose server device (hereinafter referred to as a storage node). Demand for SDS is increasing because it does not require dedicated hardware and is highly scalable. A known information processing system that uses SDS is an information processing system that combines multiple storage nodes, each of which is implemented with one or more SDS, to form a cluster and provides the cluster to a higher-level device (hereinafter referred to as a host) as a single storage device.

[0003] For example, Patent Document 1 discloses an information processing system that provides a virtual logical volume (virtual volume) from a plurality of storage nodes in which SDS is implemented. In the information processing system of Patent Document 1, storage controllers (control software) are configured as active-standby combinations across a plurality of storage nodes for redundancy, and these storage controller combinations are configured in a daisy-chain between the storage nodes. Furthermore, in the information processing system of Patent Document 1, a plurality of paths (multipaths) from a higher-level device to a storage node can be set, a first priority path is set for a storage node in which an active control software exists, and a second priority path is set for a storage node in which a standby control software exists. In the information processing system of Patent Document 1, such a multipath setting allows the active control software to switch paths to be used according to priority in the event that a failure occurs in the active control software and the standby control software is promoted to active, thereby enabling IO execution to the virtual volume to be continued, thereby realizing a redundant path configuration. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2019-185328 A Summary of the Invention [Problem to be solved by the invention]

[0005] Incidentally, in the SCSI (Small Computer System Interface) Conglomerate LUN structure, one or more SLUs (Subsidiary Logical Units) that provide logical volumes based on physical storage devices and ALUs (Administrative Logical Units) that receive IOs to the SLUs are created in a storage node, and all accesses from the host are made via the ALUs, and the ALUs distribute IOs to SLUs bound to the ALUs. In this SCSI architecture model, if the host can recognize the ALUs bound to the SLUs to be processed, it is possible to execute IOs between the SLUs via the ALUs. In this case, if the IO processing path (specifically, the path from the host side initiator to the ALU via the storage node side target and the binding from the ALU to the SLU) is a straight relationship that does not cross different storage nodes, high IO performance and high reliability due to data locality can be expected.

[0006] However, when trying to realize the information processing system of Patent Document 1 using the above-mentioned SCSI architecture model, even if the target, ALU, and SLU can be constructed in a straight relationship during normal operation, if the storage configuration is changed by adding or removing storage nodes, the straight relationship may not be maintained. If the straight relationship is broken and a connection configuration is made across storage nodes, IO processing will be performed via an external network connecting storage nodes, which will take longer to process than if IO processing were performed via the internal network of the same storage node, and IO performance will decrease. In addition, reliability will decrease if IO processing is performed via an external network.

[0007] The present invention has been made in consideration of the above points, and aims to propose an information processing system and a configuration management method that can achieve redundancy of data control in an information processing system that provides virtual volumes from the storage of multiple nodes, and that can suppress deterioration of IO performance and reliability during system operation. [Means for solving the problem]

[0008] In order to solve the above problem, the present invention provides an information processing system including a plurality of storage nodes each having one or more processors, and a storage device, the plurality of storage nodes being connected to one another via a network, , theory a first logical unit which is a logical volume; One or more A connection relationship is established between the first logical unit and the R,a second logical unit that receives an IO request from a host device for data stored in the first logical unit having the connection relationship; and a control unit that runs on the processor and processes the data stored in the first logical unit, wherein the multiple control units implemented in different storage nodes are managed as a redundancy group, and one or multiple first logical units are associated with the redundancy group, a pool based on a storage area of ​​the storage device is managed by the redundancy group, the redundancy group creates the first logical unit using the pool it manages, and the multiple control units constituting the redundancy group include an active control unit that processes the IO request to the first logical unit associated with the redundancy group, and a standby control unit that takes over the processing when a failure occurs in the active control unit, the second logical unit is disposed in a storage node in which the active control unit exists; The processor, When receiving an instruction to create the first logical unit, the active control unit is designated in accordance with a predetermined rule to create the first logical unit, and the designated The storage node in which the active control unit exists prior to an instruction to create the first logical unit. Place The above A second logical unit; Created in response to an instruction to create a first logical unit the first logical unit; The above connection relationship is set between An information processing system is provided.

[0009] In order to solve the above problem, the present invention provides a configuration management method for an information processing system including a plurality of storage nodes, each of which includes one or more processors, and a storage device, the plurality of storage nodes being connected to one another via a network, the storage nodes including: theory a first logical unit which is a logical volume; One or more A connection relationship is established between the first logical unit and the R,a second logical unit that receives an IO request from a host device for data stored in the first logical unit having the connection relationship; and a control unit that runs on the processor and processes the data stored in the first logical unit, wherein the multiple control units implemented in different storage nodes are managed as a redundancy group, and one or multiple first logical units are associated with the redundancy group, a pool based on a storage area of ​​the storage device is managed by the redundancy group, the redundancy group creates the first logical unit using the pool it manages, and the multiple control units constituting the redundancy group include an active control unit that processes the IO request to the first logical unit associated with the redundancy group, and a standby control unit that takes over the processing when a failure occurs in the active control unit, the second logical unit is disposed in a storage node in which the active control unit exists; The processor, When receiving an instruction to create the first logical unit, the active control unit is designated in accordance with a predetermined rule to create the first logical unit, and the designated The storage node in which the active control unit exists prior to an instruction to create the first logical unit. Place The above A second logical unit; Created in response to an instruction to create a first logical unit the first logical unit; The above connection relationship is set between ,A configuration management method is provided. Effect of the Invention

[0010] According to the present invention, in an information processing system that uses a plurality of storage nodes to provide virtual volumes, it is possible to realize redundant data control and suppress degradation of IO performance and reliability during system operation. [Brief description of the drawings]

[0011] [Figure 1] 1 is a block diagram showing an example of the overall configuration of an information processing system 1 according to a first embodiment of the present invention. [Diagram 2] 2 is a block diagram showing an example of a hardware configuration of a host 2. FIG. [Diagram 3] 2 is a block diagram showing an example of a hardware configuration of a storage node 3. FIG. [Figure 4]1 is a block diagram showing an example of a logical configuration of an information processing system 1. [Diagram 5] 2 is a diagram for explaining the mechanisms of ALU and SLU in the information processing system 1. FIG. [Figure 6] 2 is a block diagram showing an example of a logical configuration of a memory 32 of a storage node 3. FIG. [Figure 7] FIG. 13 illustrates an example of a system management table 64. [Figure 8] FIG. 13 is a diagram showing an example of a control software management table 65. [Figure 9] 13 illustrates an example of an ALU management table 62. [Figure 10] 13 illustrates an example of an SLU management table 63. FIG. [Figure 11] FIG. 13 shows an example of a host management table 66. [Figure 12] FIG. 13 is a diagram showing an example of a path management table 67. [Figure 13] 13 is a flowchart illustrating an example of a processing procedure for deploying a storage controller in the information processing system 1. [Figure 14] 13 is a flowchart showing detailed example processing procedures of an ALU creation process and a path setting process. [Figure 15] 13 is a flowchart showing detailed example processing procedures of an SLU creation process and a binding process; [Figure 16] FIG. 13 is a diagram illustrating an example of an SLU management table 71 in the second embodiment. [Figure 17] 13 is a flowchart illustrating an example of a processing procedure for SLU creation processing according to the second embodiment. [Figure 18] 13 is a flowchart showing an example of a processing procedure for SLU creation processing in a first modification method according to the third embodiment; [Figure 19] 13 is a flowchart illustrating an example of a processing procedure for a bind process in a second modification method according to the third embodiment. [Figure 20] 13 illustrates an example of a system operation information management table 72. FIG. [Figure 21]13 is a flowchart illustrating an example of a processing procedure for node load adjustment processing. [Figure 22] FIG. 13 is a block diagram showing an example of the configuration of an information processing system 8 according to a fifth embodiment. [Figure 23] FIG. 23 is a diagram illustrating an example of an SLU management table 91 in the fifth embodiment. [Figure 24] 23 is a flowchart illustrating an example of a processing procedure for SLU creation processing according to the fifth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0013] The following description and drawings are illustrative for explaining the present invention, and are omitted and simplified as appropriate for clarity of explanation. In addition, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention. The present invention is not limited to the embodiments, and all application examples that match the idea of ​​the present invention are included in the technical scope of the present invention. A person skilled in the art can make various additions and modifications to the present invention within the scope of the present invention. The present invention can be implemented in various other forms. Unless otherwise specified, each component may be multiple or singular. For example, several embodiments of the present invention will be described below, but in these embodiments, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, add a configuration of another embodiment to the configuration of one embodiment, or add, delete, or replace a part of the configuration of each embodiment with another configuration.

[0014] In the following explanation, various information may be explained using expressions such as "table," "list," and "queue," but the various information may be expressed in other data structures. To indicate independence from the data structure, "XX table," "XX list," and the like may be referred to as "XX information." When explaining the content of each piece of information, expressions such as "identification information," "identifier," "name," "ID," and "number" are used, but these are interchangeable.

[0015] In addition, in the following description, when describing elements of the same type without distinguishing between them, reference signs or common numbers in reference signs will be used, and when describing elements of the same type while distinguishing between them, the reference signs of those elements will be used or an ID assigned to those elements will be used instead of the reference signs.

[0016] In the following description, the process performed by executing a program may be described, but the program is executed by at least one processor (e.g., CPU) to appropriately perform a predetermined process using a storage resource (e.g., memory) and / or an interface device (e.g., communication port), and therefore the subject of the process may be the processor. Similarly, the subject of the process performed by executing a program may be a controller, device, system, computer, node, storage system, storage device, server, management computer, client, or host having a processor. The subject of the process performed by executing a program (e.g., processor) may include a hardware circuit that performs a part or all of the process. For example, the subject of the process performed by executing a program may include a hardware circuit that performs encryption and decryption, or compression and decompression. The processor operates as a functional unit that realizes a predetermined function by operating according to the program. The device and system including the processor are devices and systems that include these functional units.

[0017] A program may be installed in a device such as a computer from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium. When the program source is a program distribution server, the program distribution server includes a processor (e.g., a CPU) and a storage resource, and the storage resource may further store a distribution program and a program to be distributed. Then, the processor of the program distribution server may execute the distribution program, thereby distributing the program to be distributed to other computers. Also, in the following description, two or more programs may be realized as one program, and one program may be realized as two or more programs.

[0018] In addition, the ALU (Administrative Logical Unit) and SLU (Subsidiary Logical Unit) used in each of the following embodiments are logical units (LUs) whose specifications are defined in the Conglomerate LUN structure, which is a SCSI (Small Computer System Interface) architecture model. In the above architecture model, the SLU is a virtual logical volume (virtual volume) based on a storage device that stores actual data. In other words, the SLU is a logical unit as a logical storage area based on the physical storage area of ​​the storage device. The ALU is a volume that serves as a receiving port for access (IO request) from a host to an SLU with which a connection relationship (bind) is set with the ALU, and distributes the received IO request to the SLU. That is, in the information processing system of this description, an IO request from a host is sent to an ALU connected to an SLU that stores the target data, and the ALU that receives the IO request serves as a gateway to distribute the IO request to the above-mentioned SLU.

[0019] (1) First embodiment (1-1) System Configuration Fig. 1 is a block diagram showing an example of the overall configuration of an information processing system 1 according to a first embodiment of the present invention. The information processing system 1 shown in Fig. 1 is configured to include a plurality of hosts 2 and a plurality of storage nodes 3. Fig. 2 is a block diagram showing an example of the hardware configuration of the host 2, and Fig. 3 is a block diagram showing an example of the hardware configuration of the storage node 3.

[0020] Each host 2 and each storage node 3 are connected via a storage service network 4 consisting of, for example, Fibre Channel, Ethernet (registered trademark), InfiniBand, or a wireless LAN (Local Area Network), and each storage node 3 is connected via a back-end network 5 consisting of a LAN, Ethernet, InfiniBand, or a wireless LAN.

[0021] The storage service network 4 and the back-end network 5 may be configured as the same network, and each host 2 and each storage node 3 may be connected to a management network (not shown) other than the storage service network 4 and the back-end network 5. Furthermore, a management computer may be connected to the management network. In this case, a reference request or a setting request from the host 2 to the storage node 3 may be executed via the management network or via the management computer. Furthermore, each of the storage service network 4 and the back-end network 5 may be made redundant.

[0022] The host 2 is a physical computer device that has a function of reading and writing data to the storage node 3 via the storage service network 4 in response to a request from a user operation or an implemented application program (hereinafter referred to as an application). Note that the host 2 may be a virtual computer device such as a virtual machine (VM), and may be configured such that one or more virtual machines are constructed within the host 2.

[0023] As shown in FIG. 2, the host 2 comprises one or more CPUs (Central Processing Units) 21, one or more storage devices 23, and one or more communication devices 24 interconnected via an internal network 20, and one or more memories 22 connected to the CPU 21.

[0024] The CPU 21 is a processor that controls the overall operation of the host 2. The memory 22 is composed of volatile semiconductor memory such as SRAM (Static Random Access Memory (Random Access Memory)) and DRAM (Dynamic RAM) or non-volatile semiconductor memory, and is used as a work memory for the CPU 21.

[0025] The storage device 23 is composed of a large-capacity non-volatile storage device such as an HDD (Hard Disk Drive), SSD (Solid State Drive) or SCM (Storage Class Memory), and is used for long-term storage of various programs, control data, etc. The programs stored in the storage device 23 are loaded into the memory 22 when the host 2 is started up or when necessary, and the CPU 21 executes the programs loaded into the memory 22, thereby executing various processes of the host 2 as a whole, as described below.

[0026] The communication device 24 is an interface for the host 2 to communicate with the storage node 3 via the storage service network 4, and is composed of, for example, a Fibre Channel card, an Ethernet card, an InfiniBand card, a wireless LAN card, etc. The communication device 24 performs protocol control during communication with the storage node 3 via the storage service network 4.

[0027] The storage node 3 is a physical general-purpose server device that provides a storage area for reading and writing data to the host 2. The storage node 3 may be a virtual machine. Also, the storage node 3 may be configured to coexist with the host 2 in the same physical node.

[0028] 3, the storage node 3 includes one or more CPUs 31, a plurality of storage devices 33, one or more first communication devices 34, and one or more second communication devices 35, which are interconnected via an internal network 30, and one or more memories 32 connected to the CPU 31. Of these, the functions and configurations of the CPU 31 and memory 32 are similar to those of the corresponding parts (CPU 21 or memory 22) of the host 2, and therefore detailed explanations will be omitted.

[0029] The storage device 33 is composed of a large-capacity non-volatile storage device such as an HDD, SSD or SCM, and is connected to the second communication device 35 via an interface such as NVMe (Non-Volatile Memory Express), SAS (Serial Attached SCSI (Small Computer System Interface)), or SATA (Serial ATA (Advanced Technology Attachment)).

[0030] Moreover, the first communication device 34 is an interface through which the storage node 3 communicates with the host 2 via the storage service network 4, and the second communication device 35 is an interface through which the storage node 3 communicates with other storage nodes 3 via the backend network 5. These first and second communication devices 34, 35 have the same configuration as the communication device 24 of the host, and therefore detailed description thereof will be omitted.

[0031] In this embodiment, each storage node 3 is managed together with one or more other storage nodes 3 in a group called a cluster 6, as shown in Fig. 1. The example in Fig. 1 illustrates a case in which only one cluster 6 is set, but multiple clusters 6 may be provided in the information processing system 1. The storage nodes 3 constituting one cluster 6 are recognized by the host 2 as one storage device.

[0032] 4 is a block diagram showing an example of a logical configuration of the information processing system 1. The information processing system 1 is configured to include one or more hosts 2 and a plurality of storage nodes 3 connected to the hosts 2.

[0033] As shown in Fig. 4, an initiator IT is defined in the host 2, which is associated with a logical unit (logical unit LU in Fig. 5, described later) created in the host 2. The initiator IT is associated with any of the ports provided in the host 2. Meanwhile, a target TG is defined in each storage node 3, which is associated with each of the SLUs created in a cluster (not shown). The target TG is associated with any of the ports provided in the storage node 3.

[0034] In addition to the target TG described above, each storage node 3 is configured to include an ALU 51, one or more SLUs 52, control software 53, a cluster control unit 54, and a configuration management unit 55. The detailed mechanisms of the ALU 51 and SLU 52 will be described later with reference to Fig. 5, but the storage node 3 has an ALU 51 that receives IO requests from the host 2 and an SLU 52 that processes IO requests received by the ALU 51, and the host 2 can execute IO with the corresponding SLU 52 simply by recognizing the ALU 51. The cluster control unit 54 and configuration management unit 55 will be described later with reference to Fig. 6.

[0035] The control software 53 is a storage controller (software) that manages the storage configuration and IO (Input / Output), and is configured for redundancy with an active-standby combination across multiple storage nodes 3, as shown in Fig. 4. The active control software 53 means that it is set to a working system state (active mode) in which it can accept IO requests from the host 2, and the standby control software 53 means that it is set to a standby system state (standby mode) in which it does not accept IO requests from the host 2.

[0036] Furthermore, in the information processing system 1 according to this embodiment, redundancy groups (storage controller groups) 56 combining the above-mentioned active and standby control software 53 are configured in a daisy-chain fashion in each storage node 3. Note that in Fig. 4, one redundancy group 56 includes one active control software 53 and one standby control software 53, but in this embodiment, one redundancy group 56 may be configured with three or more control software 53 (more specifically, one active control software 53 and two or more standby control software 53).

[0037] As described above, a redundancy group (storage controller group) 56 formed from multiple control software 53 is configured in a daisy chain across multiple storage nodes 3, so that the actual data stored in SLU 52 is controlled by the control software 53 in a redundant manner between the storage nodes 3.

[0038] In this description, the connection between the initiator IT, target TG, and ALU 51 is called a path (e.g., connections 57 and 58 in FIG. 4), and the connection between the ALU 51 and SLU 52 is called a bind (e.g., connection 59 in FIG. 4). For example, in FIG. 4, as a connection when the control software 53 of storage node #1 is active, a path is set from the initiator (Initiator #1) of host #1 via the target (Target #1) of storage node #1 to ALU #1, and a bind is set from ALU #1 to SLU #1, as shown by solid lines. This connection relationship is a straight relationship because neither the path from the target TG to ALU 51 nor the bind from ALU 51 to SLU 52 crosses the storage node 3.

[0039] 4 also shows by dashed lines that a path is set from Initiator #1 via Target #2 to ALU #1, and a bind is set from ALU #1 to SLU #1. This connection relationship shows an image of the path and bind used when a failure occurs in the active control software 53 of storage node #1, and the standby control software 53 (i.e., the control software 53 of storage node #2) that forms the redundancy group 56 with that control software 53 is promoted to active.

[0040] Information on the ALU 51 and SLU 52 in the storage node 3 is always shared by the control software 53 included in the storage node 3 and the control software 53 in the redundancy group 56. Although detailed explanation is omitted, when a failure occurs in the active control software 53 and the standby control software 53 is promoted to active, the storage node 3 on which the ALU 51 and SLU 52 operate is configured to appear to operate with the promoted control software 53. As a result, to explain in detail with reference to FIG. 4, under normal circumstances, the active control software 53 on the storage node #1 appears to operate on the storage node #1 while managing the ALU #1 and SLU #2. Here, when a failure occurs in the active control software 53 and the standby control software 53 on the storage node #2 is promoted to active, a failover process is performed and the ALU #1 and SLU #1 are moved to the storage node #2, and the ALU #1 and SLU #2 appear to operate on the storage node #2. At this time, an IO request from host #1 to SLU #1 is made using the dashed connection relationship shown in Figure 4, and since Target #2, ALU #1, and SLU #1 are all located on storage node #2, the paths and binds between them are straight.

[0041] In this way, in the information processing system 1, I / O across storage nodes 3 can be executed depending on the settings of paths (connections between the initiator IT and target TG and the ALU) and binds (connections between the ALU and SLU).

[0042] Fig. 5 is a diagram for explaining the mechanisms of ALU and SLU in the information processing system 1. For ease of explanation, Fig. 5 shows only one host 2 and one storage node 3, but as shown in Fig. 1, the actual information processing system 1 may include one or more hosts 2 and multiple storage nodes 3 connected to each host 2.

[0043] As described above, the ALU 51 and SLU 52 used in this embodiment are logical units (LUs) whose specifications are defined in the Conglomerate LUN structure, which is a SCSI architecture model. The SLU 52 provides a virtual logical volume (virtual volume) based on the storage device 33 that stores actual data, and the ALU 51 serves as a receptacle for IO (Input / Output) for the SLU 52 to which it is bound. In other words, all accesses from the host 2 to the SLU 52 are made via the ALU 51, and the ALU 51 serves as a gateway that distributes IO to the bound SLU 52.

[0044] 5, in the information processing system 1, one or more SLUs 52 (SLU#1, SLU#2, ...) and an ALU 51 that receives IO requests from the host 2 are created in the storage node 3. Also, in the host 2, one or more virtual machines (VMs) 41 are constructed as virtual computers corresponding to each SLU 52. Then, the virtual volumes provided by each SLU 52 of the storage node 3 are provided to the host 2 (VM 41) from the ALU 51 via the logical unit LU.

[0045] Each VM 41 transmits an IO request to the storage node 3 in response to a user operation or a request from an implemented application program (hereinafter also referred to as an application). Here, in the information processing system 1 according to this embodiment, the binding relationship between the ALU 51 and the SLU 52 is managed in the storage node 3 by an SLU management table 63, which will be described later. Therefore, when issuing an IO request, the host 2 (VM 41) only needs to mount the ALU 51 without recognizing the SLU 52. Then, the IO request transmitted from the host 2 (VM 41) to the ALU 51 is distributed by the configuration management unit 55 to an appropriate SLU 52 in the ALU 51, and various requests such as Read / Write are transferred to the SLU 52 that is the processing target of the IO request.

[0046] Fig. 6 is a block diagram showing an example of a logical configuration of the memory 32 of the storage node 3. As shown in Fig. 6, the memory 32 of each storage node 3 stores a plurality of control software 53 (one of which is active), a cluster control unit 54, a configuration management unit 55, management information 60 for each control software 53, a system management table 64, a control software management table 65, a host management table 66, and a path management table 67.

[0047] The control software 53 is software that receives requests from the compute nodes (host 2, VM 41), performs IO processing, makes data redundant between the storage nodes 3, and stores the data in the drive (storage device 33).

[0048] The control software 53 manages management information 60 for each active-standby redundancy group (storage controller group) 56 across multiple storage nodes 3, and stores the management information 60 managed by the control software 53 of another storage node 3 belonging to the same redundancy group 56 in a synchronized manner. As shown in FIG. 6, each management information 60 has different types (active / standby) of management information 60 depending on whether the state of the control software 53 in the storage node 3 is active or standby. That is, in FIG. 6, the active management information 60 and the standby management information 60 stored in the memory 32 of the same storage node 3 are different information (the configuration of the included tables may be the same). Note that, among the management information 60, information other than the logical-physical conversion table included in the configuration information 61 may be stored in a synchronized manner not only among the control software 53 belonging to the same redundancy group 56 but also among all the storage nodes 3. However, for simplicity, in this description, a configuration example is shown in which the information is stored in a synchronized manner only within the redundancy group 56 having the active-standby relationship.

[0049] The management information 60 includes configuration information 61, an ALU management table 62, and an SLU management table 63. The configuration information 61 specifically stores management information such as a logical-physical conversion table required for storage control such as thin provisioning, hierarchical control, snapshot, compression, deduplication, or remote copy. Note that specific data examples of the configuration information 61 are publicly known, and therefore detailed explanations are omitted. The ALU management table 62 is management information for the ALU 51. A specific example is given in FIG. 9, which will be described later, to explain the ALU management table 62 in detail. The SLU management table 63 is management information for the SLU 52, and includes binding information between the ALU 51 and the SLU 52. A specific example is given in FIG. 10, which will be described later, to explain the SLU management table 63 in detail.

[0050] The cluster control unit 54 is software that manages the entire cluster (for example, alive monitoring, failover processing, etc.). The cluster control unit 54 is similar to a program that realizes a cluster control function in a known storage system, so a detailed description will be omitted.

[0051] The configuration management unit 55 is software that receives various management operations for the storage system (information processing system 1), such as referencing and changing settings, and updates management information. Note that in this embodiment, the configuration management unit 55 does not necessarily have to exist in the storage node 3, and may exist, for example, on the management computer described above.

[0052] The system management table 64 indicates information on the entire information processing system 1, and is, for example, information for managing resources owned by each storage node 3. The information in the system management table 64 is synchronized among all storage nodes 3. In FIG. 7 described later, the system management table 64 will be described in detail with a specific example.

[0053] The control software management table 65 is management information for the control software 53 arranged in each storage node 3, and the information is synchronized in all storage nodes 3. The control software management table 65 will be described in detail with a specific example in FIG.

[0054] The host management table 66 is management information for the host 2, and the information is synchronized among all the storage nodes 3. The host management table 66 will be described in detail with a specific example in FIG.

[0055] The path management table 67 is path management information, and specifically, manages information indicating an IO path from the host 2 to the ALU 51 of the storage node 3. The path management table 67 has information synchronized among all storage nodes 3. The path management table 67 will be described in detail with reference to a specific example in FIG. 12, which will be described later.

[0056] (1-2) Data structure 7 is a diagram showing an example of the system management table 64. The system management table 64 manages, for example, resources that each storage node 3 has as information on the entire information processing system 1.

[0057] 7 is configured to include a node ID 641 indicating the identifier of a storage node 3 that holds a resource to be managed (hereinafter, the resource), a resource type 642 indicating the type of the resource, a resource ID 643 indicating an identifier assigned to the resource, and a value 644 indicating the value of a predetermined item related to the resource. The predetermined item indicated by the value 644 may be predetermined according to the type of resource, and specifically, for example, if the resource type 642 is a port, a port number may be held, and if the resource type 642 is a drive, the capacity of the drive may be held, etc.

[0058] 8 is a diagram showing an example of the control software management table 65. The control software management table 65 is management information for the control software 53 arranged in each storage node 3.

[0059] The control software management table 65 shown in FIG. 8 is configured to include a storage controller ID 651 , a status 652 , an SCG ID 653 , a node ID 654 , a management capacity 655 , and a free capacity 656 .

[0060] The storage controller ID 651 indicates an identifier assigned to each control software 53 (in other words, a storage controller). The status 652 indicates the status of the control software 53 to be managed, and may be prepared, for example, not only active or standby, but also dead indicating a state in which a failure has occurred. Furthermore, not only standby indicating a reserve state, but also passive indicating a waiting state may be prepared, and other values ​​may also be used.

[0061] The SCG ID 653 indicates the storage controller group ID to which the managed control software 53 belongs. The storage controller group ID is an identifier assigned to each redundancy group (storage controller group) 56 formed including a combination of active-standby control software 53. The node ID 654 indicates the identifier of the storage node 3 in which the managed control software 53 is located. The managed capacity 655 indicates the total value of the capacity (pool) managed in units of the redundancy group (storage controller group) 56 to which the managed control software 53 belongs. The pool is used by allocating a storage area from the drive (storage device 33). The free capacity 656 indicates the capacity of the pool that can be created by the storage controller group to which the managed control software 53 belongs.

[0062] 9 is a diagram showing an example of the ALU management table 62. The ALU management table 62 is management information for the ALU 51 arranged in each storage node 3.

[0063] The ALU management table 62 shown in Figure 9 is composed of an ALU ID 621 indicating an identifier assigned to the ALU 51 to be managed, a name 622 indicating the name of the ALU 51 to be managed, and an SCG ID 623 indicating the identifier of the redundancy group (storage controller group) 56 to which the control software 53 that controls the ALU 51 to be managed belongs.

[0064] 10 is a diagram showing an example of the SLU management table 63. The SLU management table 63 is management information for the SLU 52 arranged in each storage node 3.

[0065] The SLU management table 63 shown in Figure 10 is composed of an SLU ID 631 indicating the identifier assigned to the managed SLU 52, a name 632 indicating the name of the managed SLU 52, a capacity 633 indicating the size of the logical volume provided by the managed SLU 52, an SCG ID 634 indicating, by a storage controller group ID, to which redundancy group 56 the control software 53 controlling the managed SLU 52 belongs, and an ALU ID 635 indicating the ALU ID of the ALU 51 that is in a bound state with the managed SLU 52.

[0066] When the value of the ALU ID 635 is "N / A", this means that the managed SLU 52 is not bound to any ALU 51. In the storage node 3, it is possible to create an N:M (all natural numbers) binding relationship between the SLU 52 and the ALU 51, and therefore multiple values ​​(ALU IDs) may be set in the ALU ID 635.

[0067] 11 is a diagram showing an example of the host management table 66. The host management table 66 is management information for the host 2 connected to the storage node 3 as a higher-level device.

[0068] The host management table 66 shown in FIG. 11 is configured to have a host ID 661 , a type 662 , an upper limit number of paths to be set 663 , and an initiator ID 664 .

[0069] The host ID 661 indicates an identifier assigned to the host 2. The type 662 indicates the type of the target host. Specifically, any information expressing the type can be set, such as a physical machine or virtual machine, hypervisor, or OS type. The path setting upper limit number 663 indicates the upper limit number of paths (IO routes) that can be set from the target host 2 to one LU. The initiator ID 664 indicates an identifier assigned to an initiator IT defined on the host 2 side (such as a Fibre Channel WWN (World Wide Name) or an iSCSI IQN (iSCSI Qualified Name)).

[0070] 12 is a diagram showing an example of the path management table 67. The path management table 67 is path management information, and shows information on the IO path from the host 2 to the ALU 51 of the storage node 3.

[0071] The path management table 67 shown in FIG. 12 is configured to include an ALU ID 671 , a host ID 672 , an initiator ID 673 , a node ID 674 , a target ID 675 , and a priority flag 676 .

[0072] The ALU ID 671, the initiator ID 673, and the target ID 675 respectively indicate the identifiers of the ALU 51, the initiator IT, and the target TG included in the target path. The identifier of the target TG indicated in the target ID 675 is the identifier of the port of the storage node 3 where the target TG is defined, and a value corresponding to the resource ID 643 or the value 644 (only if the value is unique) of the port in the system management table 64 in Fig. 7 is set.

[0073] The host ID 672 indicates the identifier of the host 2 in which the initiator IT included in the target path is defined, and the node ID 674 indicates the identifier of the storage node 3 having the ALU 51 and the target TG included in the target path. Note that if the port ID of the initiator IT is registered in the initiator ID 673 and this ID value is unique in the WW (World Wide), the path management table 67 may be configured not to have the host ID 672. Similarly, if the port ID of the target TG is recorded in the target ID 675 and this ID value is unique in the WW (World Wide), the path management table 67 may be configured not to have the node ID 674.

[0074] The priority flag 676 indicates information for specifying a path for processing IO preferentially when there are two or more paths (IO paths) connecting the ALU 51 indicated by the ALU ID 672 and the initiator IT indicated by the initiator ID 673. Specifically, in FIG. 12, a path for which "Yes" is registered in the priority flag 676 indicates that it is a preferred path for processing IO preferentially, and a path for which "No" is registered indicates that it is a redundant path for not processing IO preferentially. The information of the priority flag 676 is also held on the host 2 side. More generally, the priority flag 676 indicates the equivalent of Preferred in the ALUA (Asymmetric Logical Unit Access) function.

[0075] (1-3) Processing The following describes in detail the processing executed by the information processing system 1 having the above-mentioned configuration.

[0076] Fig. 13 is a flowchart showing an example of a processing procedure for deploying a storage controller in the information processing system 1. Fig. 13 shows an overview of the processing, and detailed processing procedures are shown in Figs. 14 and 15 described later.

[0077] 13 is executed when a storage controller (control software 53) is deployed (placed) in the information processing system 1, more specifically, at the time of initial construction when the active control software 53 in the information processing system 1 constructs a combination of an ALU 51 and an SLU 52 for the first time, or at the time of node addition when the configuration of the combination of an ALU 51 and an SLU 52 is changed by adding a storage node 3. Note that the information in the system management table 64 and the control software management table 65 is set based on the device configuration of the information processing system 1 before the processing in Fig. 13 is started. It is also assumed that the information in the host management table 66 is set in advance by an administrator or the like before the processing in Fig. 13 is started.

[0078] According to FIG. 13, the configuration management unit 55 first executes an ALU creation process for creating the ALU 51 (step S11).

[0079] Next, the configuration management unit 55 executes a path setting process for setting a path from the host 2 to the ALU 51 created in step S11 so as to include a port where the active or standby control software 53 exists (step S12).

[0080] Next, the configuration management unit 55 executes an SLU creation process for creating an SLU 52 in response to a request from the administrator or a higher-level program of the host 2 (step S13).

[0081] Next, the configuration management unit 55 executes a bind process to bind the ALU 51 created in step S11 and the SLU 52 created in step S13 (step S14), and then ends the overall process.

[0082] Fig. 14 is a flowchart showing an example of detailed processing procedures of the ALU creation process and the path setting process. The process shown in Fig. 14 corresponds to the ALU creation process in step S11 and the path setting process in step S12 in Fig. 13, and is specifically called and executed, for example, when a cluster environment is constructed, when a storage node 3 is added, or when the function of the ALU or SLU is enabled (when a license for enabling the function is set), etc.

[0083] 14, first, the configuration management unit 55 refers to the control software management table 65 to confirm the position of the storage node 3 in which the active control software 53 exists (step S101). Hereinafter, the control software 53 in the active state may be referred to as "active control software," and the control software 53 in the standby state may be referred to as "standby control software."

[0084] Next, the configuration management unit 55 instructs the active control software to create an ALU 51, and updates the ALU management table 62 for the created ALU 51 (step S102). The configuration management unit 55 executes the process of step S102 for all of the active control software 53 confirmed in step S101.

[0085] Next, the configuration management unit 55 checks whether or not the creation of ALUs 51 has been completed for all of the active control software 53 checked in step S101 (step S103), and if there is any incomplete control software 53 remaining (NO in step S103), returns to step S102 and repeats the creation of ALUs 51. On the other hand, if the creation of ALUs 51 has been completed for all of the active control software (YES in step S103), proceeds to step S104.

[0086] The above-mentioned processes from step S101 to when YES is determined in step S103 correspond to the ALU creation process (step S11 in FIG. 13), and the processes from step S104 onwards correspond to the path setting process (step S12 in FIG. 13).

[0087] In step S104, the configuration management unit 55 refers to the control software management table 65 and the system management table 64, and obtains port information of the storage node 3 in which the active control software exists (information linked to "Port" of resource type 642) and port information of the storage node 3 in which the standby control software that is paired with the active control software exists.

[0088] Next, the configuration management unit 55 refers to the host management table 66 to obtain the initiator ID 664 of the host 2, and sets up multiple paths from this target TG to the ALU 51 created in step S102, including a path that goes through the port (target TG) where the active control software exists, and a path that goes through the port (target TG) where the standby control software that is paired with the active control software exists (step S105).

[0089] 4, for example, when the active control software 53 existing in the storage node #1 and the standby control software 53 existing in the storage node #2 form the same redundancy group (storage controller group) 56, at least a path (active path indicated by a solid line in FIG. 4) leading to the ALU 51 via the Target #1 of the storage node #1 and a path (active path indicated by a dashed line in FIG. 4) leading to the ALU 51 via the Target #2 of the storage node #2 are set for the ALU 51 of the storage node #1 created by the active control software from the Initiator #1. In addition to the above-mentioned active path and standby path, in step S105, the configuration management unit 55 sets a path from the target TG that has acquired the initiator ID to the ALU 51 created in step S102 within the upper limit number specified in the path setting upper limit number 663 of the host management table 66.

[0090] Next, the configuration management unit 55 registers information indicating the route for each path set in step S105 in the path management table 67 (step S106). At this time, the configuration management unit 55 sets the priority flag 676 corresponding to a path going through a port of the storage node 3 where the active control software exists (the above-mentioned active path) to "Yes", thereby making the path a preferred path, and sets the priority flag 676 corresponding to a path going through a port of the storage node 3 where the standby control software exists (the above-mentioned standby path) to "No", thereby making the path a redundant path. Then, the configuration management unit 55 transmits the information in the path management table 67 after this registration to the host 2, and ends the path setting process.

[0091] 14, the path setting process (steps S104 to S106) only needs to be executed before the bind process (steps S204 to S205 in FIG. 15), and does not necessarily need to be called consecutively after the ALU creation process (steps S101 to S103) is completed. For example, the path setting process may be called when the host management table 66 is updated in response to a request from the administrator or a higher-level program of the host 2 after the ALU creation process is completed.

[0092] Furthermore, the number of ALUs 51 created in step S102 may be at least 1. In a third embodiment described later, a process for specifying the number of ALUs at the time of expansion will be described.

[0093] Fig. 15 is a flowchart showing a detailed example of the processing procedure of the SLU creation process and the binding process. The processing shown in Fig. 15 corresponds to the SLU creation process in step S13 and the binding process in step S14 in Fig. 13, and is specifically called and executed, for example, when an instruction to create an SLU is given by the administrator or a higher-level program of the host 2.

[0094] According to FIG. 15, first, the configuration management unit 55 receives a request to create an SLU from an administrator or a higher-level program of the host 2 (step S201).

[0095] Next, the configuration management unit 55 selects the control software 53 that will create the SLU 52 according to a specific rule (for example, a capacity or number of SLUs 52 defined for each control software 53 (or redundancy group) 56) (step S202). To elaborate on the specific rule in step S202, specifically, for example, a rule is adopted that preferentially selects the control software 53 with the largest free space available for creating a volume, or that selects the control software 53 with the fewest created SLUs 52. The specific rule may be preset by an administrator or a higher-level program of the host 2, or may be specified at the start of the SLU creation process, etc.

[0096] Next, the configuration management unit 55 designates the control software 53 selected in step S202 to create the SLU 52, and updates the SLU management table 63 with information about the created SLU 52 (step S203).

[0097] The above-mentioned processes from step S201 to step S203 correspond to the SLU creation process (step S13 in FIG. 13), and the processes from step S204 onwards correspond to the binding process (step S14 in FIG. 13).

[0098] In the bind process, first, the configuration management unit 55 refers to the ALU management table 62 and searches for ALUs 51 created by the same control software 53 as the SLU 52 created in step S203 (in other words, ALUs 51 controlled by the same control software 53) (step S204). If the search returns multiple ALUs 51, the ALU 51 with the fewest number of bound SLUs 52 may be selected as the ALU 51 of the search result. Alternatively, load information on each ALU 51 in the same control software 53 may be collected, and the ALU 51 with the least load may be selected as the ALU of the search result.

[0099] Next, the configuration management unit 55 executes binding between the ALU 51 found in step S204 and the SLU 52 created in step S203, updates the SLU management table 63 with information about the executed binding (step S205), and ends the binding process.

[0100] Among the processes in FIG. 15 described above, the binding process (steps S204 to S205) does not necessarily need to be called consecutively after the completion of the SLU creation process (steps S201 to S203), and may be called, for example, when a binding instruction is issued by an administrator or a higher-level program of the host 2 (such as when the virtual machine 41 is started up).

[0101] 13 to 15 are executed, and thus the information processing system 1 according to this embodiment creates at least one ALU 51 for each storage controller configured as a combination of active-standby control software 53 across storage nodes 3 when deploying the storage controller (at the time of initial construction or when adding a node). The information processing system 1 configured in this manner realizes redundancy of data control and can suppress degradation of IO performance and reliability during system operation.

[0102] 13 to 15 are executed in the information processing system 1 according to this embodiment, in the storage node 3 (storage node #1) in which the active control software 53 exists, both the path between the target TG and the ALU 51 and the bind between the ALU 51 and the SLU 52 can be set in a straight relationship that does not span across other storage nodes 3. Furthermore, in the storage node 3 (storage node #2) in which the standby control software 53 that pairs with the active control software 53 exists, a path to the ALU 51 is set via the port (target TG) of the node. By setting this standby path, when a failure occurs in the storage node #1, the standby control software of the storage node #2 is promoted to active, and the ALU 51 and SLU 52 of the storage node #1 are moved to the storage node #2 by failover, both the path and the bind become straight relationships in the storage node #2.

[0103] That is, according to the information processing system 1 of this embodiment, the storage controller is configured with a combination of active-standby control software 53, thereby realizing redundant data control. Not only can high IO performance and reliability be obtained by the straight path and bind configuration in storage node #1 during normal configuration, but even when a configuration change is made due to a failover or other such event due to a failure, straight paths and binds are configured within storage node #2 after the change, thereby maintaining high IO performance and reliability.

[0104] (2) Second embodiment The information processing system 1 according to the second embodiment will be described, focusing on the differences from the first embodiment. Repetitive description of the configuration and processing common to the first embodiment will be omitted.

[0105] The information processing system 1 according to the second embodiment executes the SLU creation process in a different procedure among the four main processes (ALU creation process, path setting process, SLU creation process, and bind process) constituting the overall process shown in FIG. 13 in the first embodiment. The details will be described later, but the SLU creation process in the second embodiment is characterized in that when creating a new SLU, if the purpose of use of the SLU is the same as that of an already created SLU, the new SLU is created in the same storage controller group as the already created SLU. Also, in the second embodiment, when creating the new SLU, if the SLU cannot be created in the same storage controller group as the already created SLU due to a problem of free space, the SLU that can be moved from among the already created SLUs is moved to another storage node 3 to increase the free space, and then the new SLU is created.

[0106] To realize the second embodiment, the information processing system 1 of this embodiment stores an SLU management table 71 having a data configuration different from that of the first embodiment in the memory 32 of the storage node 3, and executes the SLU creation process using a processing procedure different from that of the first embodiment.

[0107] Fig. 16 is a diagram showing an example of an SLU management table 71 in the second embodiment. The SLU management table 71 in Fig. 16 differs from the SLU management table 63 shown in Fig. 10 in the first embodiment in that an APP ID 716 has been added, and the information indicated by the other data items (SLU ID 711, name 712, capacity 713, SCG ID 714, and ALU ID 715) is the same as the information indicated by the data items with the same names in the SLU management table 63, and therefore detailed description thereof will be omitted.

[0108] The APP ID 716 is an identifier indicating which application (for example, the VM 41 of the host 2) uses the managed SLU 52, and indicates an identifier (application ID) given to each application. In other words, the APP ID 716 is information specifying the application that accesses and runs on the managed SLU 52. The information registered in the APP ID 716 is provided by an administrator or a higher-level program of the host 2 when the SLU 52 is created. Note that, in the SLU management table 71 of FIG. 16, the purpose of use of the SLU 52 is identified by the application ID, but the SLU management table 71 of this embodiment only needs to hold information indicating the purpose of use of the SLU 52, and may hold information indicating the purpose of use in another format (for example, database log use) without being limited to the application ID.

[0109] 17 is a flowchart showing an example of the processing procedure of SLU creation processing according to the second embodiment. Like the SLU creation processing according to the first embodiment, the SLU creation processing shown in FIG.

[0110] 17, first, configuration management unit 55 receives a request to create an SLU 52 from an administrator or a higher-level program of host 2 (step S301). This SLU creation request specifies the application ID of the application that will use the SLU 52 to be created, and the capacity of that SLU. Hereinafter, the SLU 52 whose creation is requested in step S301 will be referred to as a new SLU.

[0111] Next, the configuration management unit 55 refers to the SLU management table 71 and determines whether an SLU 52 has already been created that has the same purpose of use as the new SLU (step S302). Hereinafter, the created SLU 52 is referred to as an existing SLU. Specifically, the configuration management unit 55 determines whether there is an existing SLU in which an application ID corresponding to the application that uses the new SLU is registered in the APP ID 716 of the SLU management table 71. For example, in the SLU management table 71 of Fig. 16, if the new SLU is used by an application with application ID "1", then SLU#1 and SLU#2 of the existing SLUs SLU#1 to SLU#3 have application ID "1" registered in the APP ID 716, and therefore correspond to existing SLUs that have the same purpose of use as the new SLU.

[0112] In step S302, if there is an existing SLU that has the same purpose of use (application ID) as the new SLU (YES in step S302), proceed to step S303; if there is no existing SLU that has the same purpose of use (application ID) as the new SLU (NO in step S302), proceed to step S307.

[0113] In step S303, the configuration management unit 55 determines whether the free space for volume creation in the control software 53 that controls the existing SLU having the same purpose of use (application ID) as the new SLU is greater than the capacity required for the new SLU, and if a positive result is obtained (YES in step S303), the process proceeds to step S306; if a negative result is obtained (NO in step S303), the process proceeds to step S304.

[0114] If a negative result is obtained in step S303, the configuration management unit 55 determines whether or not there is a group of SLUs (meaning one or more SLUs) that can be moved to another storage node 3 (under the control of another control software 53) in the storage node 3 where an existing SLU having the same usage purpose as described above exists (step S304).

[0115] More specifically, the process of step S304 may be executed as follows. First, the configuration management unit 55 refers to the SLU management table 71 and searches for existing SLUs that satisfy the following conditions: ALU ID 715 is empty (N / A) (in other words, unbound) and APP ID 716 is a common application ID (not necessarily the same application ID as the new SLU), thereby identifying a group of SLUs that can be moved. Next, the configuration management unit 55 refers to the control software management table 65 and compares the total capacity of the group of SLUs that can be moved identified in the previous search with the free capacity of each of the other control software 53. Then, if there is "another control software 53" that has free capacity larger than the total capacity of the group of SLUs that can be moved, the configuration management unit 55 determines in step S304 that there is a group of SLUs that can be moved under the control of the "other control software 53".

[0116] Then, if it is determined in step S304 that there is a group of SLUs that can be moved under the control of another control software 53 (YES in step S304), the configuration management unit 55 moves the SLU group under the control of the other control software 53 (step S305). As a result of the processing in step S305, the control software 53 in which the SLU group existed before the move has the free space required for a new SLU, so the process proceeds to step S306. On the other hand, if it is determined in step S304 that there is no group of SLUs that can be moved under the control of another control software 53 (NO in step S304), the process proceeds to step S307.

[0117] The process of step S306 is executed when an existing SLU exists that has the same usage purpose as the new SLU, and the control software 53 that controls this existing SLU has the free space required to create a new SLU. At this time, the configuration management unit 55 selects the above-mentioned control software 53, i.e., the control software 53 that controls the existing SLU with the same usage purpose (for example, when performing the process of step S306 after step S305, the control software 53 from which the SLU group is moved) as the creation destination of the new SLU (step S306).

[0118] On the other hand, the process of step S307 is executed when there is no existing SLU with the same usage purpose as the new SLU, or when there is an existing SLU with the same usage purpose as the new SLU but the control software 53 that controls this existing SLU cannot secure the free space required to create a new SLU. At this time, the configuration management unit 55 selects another control software 53 in which to create the new SLU according to a specific rule (for example, the capacity or number of SLUs 52 determined for each control software 53) (step S307). The process of selecting the control software 53 in which to create the new SLU in step S307 is the same as the process of step S202 in FIG. 15.

[0119] Then, the configuration management unit 55 designates the control software 53 selected in step S306 or step S307 to create a new SLU, and updates the SLU management table 63 with information about the created SLU 52 (step S308). This ends the SLU creation process.

[0120] In step S307, if a new SLU cannot be created by the control software 53 that controls an existing SLU with the same usage purpose, the destination for creating the new SLU is selected from other control software 53 in accordance with specific rules. However, as a variant, a warning may be returned and the above processing may be executed, or an error may be returned and the SLU creation processing may be aborted.

[0121] Furthermore, the processes of steps S304 and S305 do not necessarily have to be executed in the series of processes in the SLU creation process of Fig. 17. This is because the processes of steps S304 and S305 require a relatively long processing time, and the configuration management unit 55 may execute these processes periodically without synchronizing them with the SLU creation process.

[0122] As explained above with reference to Figure 17, the SLU creation process in the second embodiment differs from the SLU creation process shown in Figure 15 in the first embodiment in that, when selecting the control software 53 to create an SLU in, the control software 53 that controls an existing SLU having the same purpose of use (application ID) as the new SLU to be created is selected first, and if there is insufficient free space in that control software 53, the existing SLU having the same purpose of use is moved to another control software 53 and then that control software 53 is caused to create a new SLU.

[0123] According to the second embodiment, multiple SLUs 52 having the same purpose of use (application ID), in other words, multiple SLUs 52 constituting the same application, can be placed, as far as possible, under the control of the same control software 53 in a single redundancy group 56 (or, more specifically, a single storage node 3).

[0124] Here, in a known virtual volume providing function, there may be multiple SLUs constituting one VM 41 (each having a different type of virtual volume). If these multiple SLUs are distributed across multiple storage nodes 3, there are disadvantages in terms of IO access performance and availability, so it is preferable to consolidate them on the same node as much as possible. For example, a disadvantage in terms of availability is that if the SLUs constituting one VM 41 are distributed across multiple storage nodes 3, the probability of node failure occurring in the SLUs constituting one VM 41 increases by the amount of node distribution.

[0125] Therefore, in the second embodiment, by executing the above-mentioned SLU creation process, multiple SLUs 52 having the same purpose of use (application ID), i.e., multiple SLUs 52 used by the same VM 41, are configured to be controlled by the same control software 53 in one storage node 3, so that in addition to the effects obtained by the first embodiment, it is possible to obtain the effect of suppressing deterioration in the access performance and availability of the above-mentioned IO.

[0126] (3) Third embodiment The information processing system 1 according to the third embodiment will be described, focusing on the differences from the first embodiment. Repetitive description of the configuration and processing common to the first embodiment will be omitted.

[0127] The information processing system 1 according to the third embodiment takes into consideration the reduction of storage nodes 3, and of the four main processes (ALU creation process, path setting process, SLU creation process, and bind process) that constitute the overall processing shown in FIG. 13 in the first embodiment, the SLU creation process and the bind process are executed using different processing procedures.

[0128] Generally, when a storage node (hereinafter, abbreviated as a node) is removed from a cluster in a storage system, the resources present in the node to be removed are moved to another node before the node is removed. Similarly, when a node is removed in the information processing system 1 according to the first embodiment, the ALU and SLU assigned to the active control software are moved to another control software, and then the node in which the active control software was present is removed from the cluster. Here, in the first embodiment, the path and bind are set to straight in the node in which the active control software is present, and if the ALU and SLU are to be migrated in this state without stopping IO, they will be moved to another control software without changing the bind setting between the ALU and SLU. However, when the number of ALUs created in a node is small (for example, one), when moving to another control software while maintaining the bind relationship between the ALU and SLU, the ALU and SLU chunks are very large, so that the capacity of a specific control software is excessively consumed, and the efficiency of distributing to another control software is reduced. This is expected to be a problem. In order to solve the above problems, the third embodiment is characterized in that, by executing the SLU creation process and binding process of this embodiment when removing a storage node 3, it is possible to change (mainly increase) the number of ALUs 51 in a storage node 3 during system operation.

[0129] In the third embodiment, there are three methods for changing the number of ALUs 51 during system operation: when an SLU is created (first change method), when a bind is set (second change method), and when a node is added (third change method). The first to third change methods are each explained below. Note that, although the following explanation focuses on increasing the number of ALUs 51, the same can be applied to decreasing the number of ALUs 51.

[0130] (3-1) First change method 18 is a flowchart showing an example of the processing procedure of SLU creation processing in the first modification method of embodiment 3. Like the SLU creation processing in the first embodiment, the SLU creation processing shown in FIG.

[0131] 18, first, the configuration management unit 55 receives a request to create an SLU 52 from an administrator or a higher-level program of the host 2 (step S401). Hereinafter, the SLU 52 for which creation has been requested is referred to as a new SLU.

[0132] Next, the configuration management unit 55 refers to the control software management table 65 and the SLU management table 63 to check whether there is any control software 53 for which the number of SLUs 52 is less than a predetermined number (more specifically, whether there is any active control software 53 controlling a predetermined number or less of SLUs 52) (step S402).

[0133] To explain the process of step S402 in detail, the configuration management unit 55 determines the storage controller group to which each SLU belongs based on the SLU ID 631 and SCG ID 634 in the SLU management table 63, and further determines the control software 53 in the active state in each storage controller group based on the storage controller ID 651, state 652, and SCG ID 653 in the control software management table 65. As a result, the configuration management unit 55 can recognize which control software 53 each SLU 52 belongs to, and can recognize the number of SLUs 52 controlled by each control software 53 and can determine whether the number is equal to or less than the "predetermined number". The "predetermined number" may be a value determined in advance in the information processing system 1, or may be a value designated by an administrator or a higher-level program of the host 2 when creating an SLU, etc., and dynamically changeable during operation.

[0134] Furthermore, in step S402 shown in FIG. 18, the determination criterion is simply the number of SLUs 52. However, as a modified example, the determination criterion may be the number of SLUs 52 that have been bound to the ALU 51.

[0135] If control software 53 with the number of SLUs 52 not exceeding the predetermined number exists in step S402 (YES in step S402), the process proceeds to step S403, where the same process as the SLU creation process in the first embodiment (specifically, steps S202 to S203 in FIG. 15) is performed. On the other hand, if control software 53 with the number of SLUs 52 not exceeding the predetermined number exists in step S402 (NO in step S402), the process proceeds to step S404.

[0136] In step S403, the configuration management unit 55 selects the control software 53 that will create the SLU 52 according to a specific rule (for example, the capacity or the number of SLUs 52 determined for each control software 53). The process of selecting the destination control software 53 in step S403 is the same as the process in step S202 in Fig. 15. After step S403 is completed, the process proceeds to step S406.

[0137] In step S404, the configuration management unit 55 searches for the control software 53 with the fewest number of ALUs 51 it controls among the control software 53 with free space larger than the capacity of the new SLU, and selects the corresponding control software 53 as the destination for creating the new SLU.

[0138] Next, the configuration management unit 55 instructs the control software 53 selected in step S404 to create an ALU 51 and set a path between the created ALU 51 and the host 2 (step S405). Note that the path setting process in step S405 is executed by the same process as steps S104 to S106 in Fig. 14. After step S405 ends, the process proceeds to step S406.

[0139] Finally, the configuration management unit 55 specifies the control software 53 selected in step S403 or step S404 to create a new SLU, updates the SLU management table 63 with information about the created SLU 52 (step S406), and terminates the SLU creation process.

[0140] As described above, according to the first modification method, in the SLU creation process, the configuration management unit 55 causes the control software 53 with the fewest SLUs to create an SLU based on the number of SLUs and / or ALUs of each control software 53 (in other words, each storage controller group), or, if there is no control software 53 with a few SLUs, causes the control software 53 with the fewest ALUs to add an ALU that is directly bound to the newly created SLU and then create a new SLU. That is, in the first modification method, the storage node 3 creates additional ALUs based on an increase in the number of SLUs due to SLU creation during system operation.

[0141] (3-2) Second change method 19 is a flowchart showing an example of a process procedure of a bind process in the second modification method of the third embodiment. As in the bind process in the first embodiment, the bind process shown in FIG.

[0142] 19, first, the configuration management unit 55 receives a bind execution request from an administrator or a higher-level program of the host 2 (step S501). The bind execution request specifies the SLU 52 to be bound.

[0143] Next, the configuration management unit 55 refers to the ALU management table 62 to search for an ALU 51 that is controlled by the same control software 53 as the SLU 52 specified in step S501 (step S502).

[0144] Next, the configuration management unit 55 refers to the SLU management table 63 and determines whether or not there is an ALU 51 for which the number of bound SLUs 52 is equal to or less than a predetermined number (step S503). Note that the "predetermined number" may be a value determined in advance in the information processing system 1, or may be a value that can be dynamically changed during operation, specified by an administrator or a higher-level program of the host 2, for example, when the bind process or the SLU creation process is executed.

[0145] If a positive result is obtained in step S503 (YES in step S503), the configuration management unit 55 executes binding between the ALU 51 found in step S502 and the SLU 52 specified in step S501, as in the bind processing in the first embodiment (specifically, step S205 in FIG. 15), updates the SLU management table 63 with information about the executed bind (step S504), and terminates the bind processing. On the other hand, if a negative result is obtained in step S503 (NO in step S503), the process proceeds to step S505.

[0146] In step S505, the configuration management unit 55 instructs the control software 53 that is the same as the SLU 52 specified in step S501 to create an ALU 51 and set a path between the created ALU 51 and the host 2 (step S505). Note that the path setting process in step S505 is executed by the same processes as steps S104 to S106 in FIG.

[0147] Next, the configuration management unit 55 executes binding between the ALU 51 created in step S505 and the SLU 52 specified in step S501, updates the SLU management table 63 with information about the executed binding (step S506), and terminates the binding process.

[0148] As described above, in the second modification method, the storage node 3 is characterized by creating additional ALUs based on the current number of binds between ALUs and SLUs (the number of connection relationships set between ALUs and SLUs) during system operation.

[0149] (3-3) The third change method The third change method is a method for changing the number of ALUs 51 when adding a storage node 3 (hereinafter, node), and a specific example of the procedure will be described below.

[0150] When a node is to be added, the configuration management unit 55 creates an additional ALU 51 by executing step S102 of the ALU creation process in Fig. 14. At this time, as a third change method, the configuration management unit 55 first refers to the control software management table 65 and the ALU management table 62 to check the number of ALUs 51 for not only the control software 53 operating on the node to be added, but also for the control software 53 operating on all nodes in the cluster.

[0151] Next, when the number of nodes constituting a cluster is N and the total number of ALUs 51 in the cluster is less than (N-1), the configuration management unit 55 causes the control software 53 to create an ALU so that the total number of ALUs 51 in each node in the cluster becomes (N-1), performs binding processing between the created ALU and at least any SLU 52 in the same node, and further causes the configuration management unit 55 to set a path between the created ALU 51 and the host 2. Note that the number of nodes, N, can also be rephrased as the number of control software 53 that are active in the cluster. The reason for creating ALUs 51 so that the number becomes (N-1) in each node in the cluster is that, when a node is removed, there are (N-1) nodes other than the node to be removed, and (N-1) destinations for the pairs of ALUs 51 and SLUs 52 are prepared to match this number of nodes (N-1).

[0152] That is, in the third modification method, the storage node 3 additionally creates ALUs 51 when a node is added so that the total number of ALUs 51 in each node becomes (N-1) for the number of nodes N after the addition.

[0153] As a result of creating (N-1) ALUs 51 in each node in the cluster as described above, when removing one of the N nodes in the cluster, it is possible to distribute one or more SLUs 52 (SLU groups) bound to each of the (N-1) ALUs 51 in the node to be removed across the (N-1) nodes that are not to be removed.

[0154] Here, the specific method of distributing SLUs 52 to other nodes when a node is removed is not limited to a specific one. For example, a conceivable method is to check the capacity of one or more SLUs 52 (SLU groups) bound to each ALU 51 in the node to be removed, and move the SLU groups with the largest total capacity value to a node with the largest free capacity among the (N-1) other nodes that are not the target of removal (this can also be rephrased as the control software 53 in the active state). In this case, more specifically, the SLU group with the largest total capacity value is moved to the other node with the largest free capacity, and the SLU group with the next largest total capacity value is moved to the remaining other node with the next largest free capacity. In this way, the SLU groups paired with each ALU 51 in the node to be removed are moved in order of the largest total capacity value to the other node with the largest free capacity. Then, in the destination node, the SLU groups moved by the control software 53 in the active state are bound to the ALU 51 of the node. According to such a distribution method, it is possible to suppress bias in the free capacity of all the nodes in the cluster after removal, and to avoid saturation of capacity.

[0155] As another example of a distribution method, one or more SLUs 52 (SLU groups) bound to ALUs 51 in the node to be removed may be distributed so as to reduce bias in performance (e.g., resource load) in the destination node due to the movement of the SLU groups. Specifically, for example, the loads on the CPUs 31 and SLUs 52 in each node including the node to be removed may be checked, and the SLU groups may be moved in order starting with the SLU groups with the highest loads in the node to be removed to nodes with the lowest CPU 31 loads among the remaining nodes in the cluster. Such a distribution method can suppress bias in the load across all the nodes in the cluster after the removal, and is expected to improve the performance of the entire node.

[0156] As described above, in the third embodiment, by adopting at least any one of the first to third change methods or a combination of these, whereas in the first embodiment at least one ALU 51 is created for each storage controller (combination of active-standby control software 53), it is possible to change the number of ALUs 51 in each storage node 3 during system operation and set a path to the host 2 for the additionally created ALU 51.

[0157] In the information processing system 1 in which an ALU 51 is additionally created as described above, the combination of ALU 51 and SLU 52 can be changed to multiple combinations. Therefore, even when the ALU 51 and SLU 52 of the node to be removed are moved without stopping IO when a node is removed, the set of ALU 51 and SLU 52 can be efficiently moved and distributed to one or more different control software 53 while enabling straight path and bind connections at the destination node.

[0158] (4) Fourth embodiment The information processing system 1 according to the fourth embodiment will be described, focusing on the differences from the first embodiment. Repetitive description of the configuration and processing common to the first embodiment will be omitted.

[0159] The information processing system 1 according to the fourth embodiment takes into consideration the performance degradation that occurs during resource migration in the information processing systems 1 according to the first to third embodiments, and is characterized in that when a storage node 3 (hereinafter referred to as a node) is removed or resources are migrated between nodes, a node load adjustment process described below is executed to consolidate the binding relationships of SLUs 52 with high loads to the same ALU 51 and migrate them to another node.

[0160] The information processing system 1 according to the fourth embodiment holds, for example, a system operation information management table 72 in the memory 32 of the node in addition to each information table described in the first embodiment.

[0161] Fig. 20 is a diagram showing an example of the system operation information management table 72. The system operation information management table 72 is data for managing load information for various resources owned by each storage node 3 of the information processing system 1. The system operation information management table 72 shown in Fig. 20 is configured to include a node ID 721, a resource type 722, a resource ID 723, a metric 724, a time 725, and an amount 726.

[0162] The node ID 721 indicates an identifier of the storage node 3 having the target resource. The resource type 722 indicates the type of the target resource. The resources recorded in the resource type 722 are not limited to hardware resources of the storage node 3 such as the CPU 31, the ports of the communication devices 34 and 35, and the drives of the storage device 33, but may also be software resources such as the ALU 51 and the SLU 52. The resource ID 723 indicates an identifier for identifying the target resource. The metric 724 indicates a metric of the load information. The time 725 indicates the measurement time of the load information. The amount 726 indicates the measured load amount. The load amount in the ALU 51 and the SLU 52 is, for example, a value indicating the access frequency, such as the IOPS of read and write.

[0163] The information managed in the system operation information management table 72 need only include information required for the node load adjustment process described below or information that can be used to calculate that information, and is not limited to the specific example shown in Figure 20.

[0164] In addition, in the system operation information management table 72 of Figure 20, information regarding ALU 51 and SLU 52 is represented in a single table by referencing the information of each SCG ID in the control software management table 65, ALU management table 62, and SLU management table 63 and linking them to the same node, but operation information (load information) regarding ALU 51 and SLU 52 may be managed in a table separate from the system operation information management table 72.

[0165] Fig. 21 is a flowchart showing an example of a processing procedure of node load adjustment processing. The node load adjustment processing is executed by the configuration management unit 55 at a predetermined timing during system operation, such as when a node is removed or when resources are transferred between nodes. The execution timing of the node load adjustment processing is assumed to be at least after step S14 in Fig. 13 or the bind processing shown in Fig. 19.

[0166] According to FIG. 21, first, the configuration management unit 55 refers to the system operation information management table 72 to check the loads of the CPU 31 and the SLU 52 in each node (step S601).

[0167] Next, the configuration management unit 55 checks whether an SLU 52 whose load exceeds a reference value and an SLU 52 whose load is equal to or less than the reference value are bound to the same ALU 51 (step S602). The "reference value" for an SLU 52 may be a value determined in advance, or a threshold value for classifying the load into high and low loads may be determined during operation. The "reference value" may be a value specified by an administrator or a higher-level program of the host 2, or may be determined by a predetermined program in the node (for example, the configuration management unit 55) calculating the amount of load that the destination node can accept based on the load information of the destination node. The reference value for determining a high load and the reference value for determining a low load may be different values ​​that do not have an overlapping section.

[0168] If a positive result is obtained in step S602 (YES in step S602), the process proceeds to step S603, and if a negative result is obtained (NO in step S602), the process proceeds to step S605, which will be described later.

[0169] In step S603, the configuration management unit 55 refers to the system operation information management table 72 and the SLU management table 63, and searches for an ALU 51 to which only an SLU 52 whose load is below the reference value is bound.

[0170] Next, the configuration management unit 55 rebinds the SLUs 52 that fall below the reference value and that met the judgment criteria in step S602 (SLUs 52 that are equal to or less than the reference value) to the ALUs 51 that matched the search in step S603 (step S604). Rebinding is a process that successively executes binding to a new connection destination and unbinding from the current binding, and is mainly used when changing the ALUs 51 to be bound to an SLU 52. As a result of the process in step S604, only SLUs 52 whose load is below the reference value are bound to the ALUs 51 to be rebound. And only SLUs 52 whose load is above the reference value are bound to the ALUs 51 that were not rebound. After step S604 is completed, the process proceeds to step S605.

[0171] In step S605, the configuration management unit 55 migrates the block of ALUs 51 to which only SLUs 52 whose load is below the reference value are bound, and the block of SLUs 52 bound to the ALUs 51, to a node with a higher load.

[0172] Next, the configuration management unit 55 migrates the block of ALUs 51 to which only SLUs 52 whose load exceeds the reference value are bound, and the SLUs 52 bound to those ALUs 51, to a node with a lower load (step S606), and terminates the node load adjustment process.

[0173] The migration destination of the node in step S605 may be selected from among the nodes with relatively high loads among all the nodes, and further, ALU 51 to which only SLU 52 with a load below a reference value is bound may be moved to the node group with relatively high load in a distributed manner according to a predetermined method.Similarly, the migration destination of the node in step S606 may be selected from among the nodes with relatively low loads among all the nodes, and further, ALU 51 to which only SLU 52 with a load above a reference value is bound may be moved to the node group with relatively low load in a distributed manner according to a predetermined method.

[0174] As described above, in the information processing system 1 according to the fourth embodiment, when a node is removed or resources are transferred between nodes, node load adjustment processing is executed to bind multiple SLUs 52 to separate ALUs 51 based on the level of the load, and the bound ALUs 51 and SLUs 52 groups are moved to different nodes, thereby making it possible to suppress uneven load distribution across the nodes. As a result, in the information processing system 1, performance variations across nodes are homogenized when resources are transferred, and improved performance across the nodes can be expected.

[0175] In the above explanation, the SLUs 52 with high loads are migrated together to another node in the node load adjustment process, but as a modification of this embodiment, the load (e.g., CPU utilization rate) of the destination node is checked, and only the SLUs 52 with performance values ​​that can withstand the node load after migration (only the total load value may be considered, not the level of the load) are migrated together into the same ALU 51, and this group is then migrated to the destination node. In such a modification, the SLUs 52 can reliably demonstrate a predetermined performance value even after the node migration.

[0176] (5) Fifth embodiment The information processing system 1 of the above-mentioned first to fourth embodiments has a configuration in which the host 2 having the upper program (e.g., VM 41) that outputs IO requests is separated from the storage node 3 that is the target of the IO requests. However, one of the important features of the present invention is the configuration management of the ALU and SLU, and the scope of its applicability is not limited to the information processing system with the above-mentioned separated configuration.

[0177] Therefore, in the fifth embodiment, an information processing system 8 having a configuration of HCI (HyperConverged Infrastructure) in which a host (or VM) and storage are arranged in the same node will be described as an example in which the present invention is applied to an information processing system having a device configuration different from that of the information processing system 1 of the first to fourth embodiments. Note that the fifth embodiment will be described mainly focusing on the differences from the first to fourth embodiments, and components and processes common to the first to fourth embodiments will be denoted by common reference numerals, and repeated description will be omitted.

[0178] 22 is a block diagram showing a configuration example of an information processing system 8 according to the fifth embodiment. The information processing system 8 is an information processing system having an HCI configuration, and is composed of a plurality of hypervisor nodes 80 connected via a network outside the nodes.

[0179] HCI is an integrated system that consolidates the functions of a host node and a storage node in one physical computer node (hypervisor node 80 in the case of FIG. 22). HCI makes it possible to execute multiple processes on one computer node by running applications, middleware, management software, containers, etc. in addition to storage software on the OS or hypervisor installed on each computer node.

[0180] In the case of the information processing system 8 shown in FIG. 22, each hypervisor node 80 is configured by coexisting a user APP VM 82 on a base hypervisor, which corresponds to the host 2 in the first to fourth embodiments and on which a user application (user APP 84) runs, and a storage OS VM 83 on which various programs (control software 53, cluster control unit 54, and configuration management unit 55) of the storage node 3 in the first, third, and fourth embodiments run. In the information processing system 8, a plurality of such hypervisors are arranged to form a cluster. In addition, each hypervisor node 80 includes a hypervisor management unit 81 that manages the IO and configuration of the VMs 82 and 83 running on the hypervisor. The hypervisor management unit 81 is a program having a general hypervisor management function, so a detailed description of the operation will be omitted.

[0181] The hypervisor node 80 has an initiator IT (e.g., Initiator #1) and a target TG (e.g., Target #1) that are associated with a physical port of the hypervisor node 80, and a virtual initiator IT (e.g., Initiator #11) and a target TG (e.g., Target #11) that are associated with a virtual port of the VMs 82 and 83. In setting a path that connects the VM 82 for user APP and the ALU 51 on the VM 83 for storage OS, the path is set using the port of the virtual initiator IT and the port of the virtual target TG.

[0182] Various tables (similar to the various tables shown in FIG. 6) managed in the memory in the VM 83 for storage OS manage the above virtual information. The SLU management table requires a different configuration from the first to fourth embodiments, which will be described later with reference to FIG. 23. Various programs (control software 53, cluster control unit 54, and configuration management unit 55) held in the memory in the VM 83 for storage OS basically execute the same operation control as in the first, third, and fourth embodiments. The difference between the second embodiment and the fifth embodiment will be described later with reference to FIG. 24.

[0183] Fig. 23 is a diagram showing an example of an SLU management table 91 in the fifth embodiment. The SLU management table 91 in Fig. 23 differs from the SLU management table 71 shown in Fig. 16 in the second embodiment in that an APP node ID 917 is added, and the information indicated by the other data items (SLU ID 911, name 912, capacity 913, SCG ID 914, ALU ID 915, and APP ID 916) is the same as the information indicated by the data items of the same names in the SLU management table 71. For example, the APP ID 916 is information specifying a user APP 84 that accesses and operates the SLU 52 to be managed.

[0184] The APP node ID 917 is an identifier indicating a node (hypervisor node 80) in which a user application (user APP 84) that uses the SLU 52 to be managed exists, and is indicated by a node ID assigned to each hypervisor node 80. The APP node ID 917 is information used to create the SLU 52 on the control software 53 that exists in the same node (hypervisor node 80) as the user APP VM 82 in the SLU creation process shown in Fig. 24, and is specified when an administrator or user APP 84 requests SLU creation.

[0185] In the second embodiment described above, the SLUs 52 used by the same application are created on the same control software 53. However, in the information processing system 8 according to the fifth embodiment having an HCI configuration, the SLUs 52 used by the same application (user APP 84) are created on the same control software 53 in the same node as the application, so that the same or better effects as those of the second embodiment can be expected. That is, when the storage OS VM 83 in which the SLU 52 exists and the user APP VM 82 using the SLU 52 are located across nodes, there is a risk of a decrease in IO performance because IO processing is performed via an external network. However, when the storage OS VM 83 in which the SLU 52 exists and the user APP VM 82 using the SLU 52 are located on the same node, IO processing is possible via the network within the hypervisor, so high IO performance can be obtained. Therefore, in the fifth embodiment, all the SLUs 52 constituting the user APP VM 82 are created in the control software 53 in the same node as the user APP VM 82 by the SLU creation process shown in FIG. 24.

[0186] Fig. 24 is a flowchart showing an example of a processing procedure of SLU creation processing according to embodiment 5. Like the SLU creation processing (see Fig. 17) according to embodiment 2, the SLU creation processing shown in Fig. 24 is executed by the configuration management unit 55.

[0187] The SLU creation process in the fifth embodiment shown in Fig. 24 differs from the SLU creation process in the second embodiment shown in Fig. 17 mainly in steps S701, S703, and S704. These different processes will be described below.

[0188] In step S701, the configuration management unit 55 receives a request to create an SLU 52 from an administrator or a higher-level program of the user APP VM 82. At this time, the SLU creation request specifies the capacity of the new SLU, the ID of the application that uses the new SLU (APP ID), and the ID of the node where the application exists (APP node ID).

[0189] Step S703 is a process performed when it is determined in step S702 that an SLU 52 (existing SLU) having the same purpose of use as the new SLU has already been created, and the configuration management unit 55 provisionally selects, as the destination for creating the new SLU, a control software 53 in which an existing SLU with the same APP ID specified in step S701 exists. In step S703, the control software 53 to be provisionally selected is determined based on the APP ID, but the same result will be obtained if provisional selection is made based on the APP node ID.

[0190] Step S704 is processing that is performed when it is determined in step S702 that an SLU 52 (existing SLU) having the same usage purpose as the new SLU has not been created, and the configuration management unit 55 provisionally selects the active control software 53 in the node (hypervisor node 80) indicated by the APP node ID specified in step S701 as the destination for creating the new SLU.

[0191] The processing of steps S703 and S704 above is for attempting whether a new SLU can be created at the node indicated by the APP node ID, and in the fifth embodiment, whenever possible, a control software 53 operating at the node specified by the APP node ID is selected as the destination for creating the new SLU.

[0192] The process from step S705 onwards is the same as the process from step S303 onwards in Fig. 17. To explain the outline, if the control software 53 provisionally selected in step S703 or step S704 has free space capable of creating a new SLU, the new SLU is created by that control software 53 (steps YES in S705 to S708 to S710), if the provisionally selected control software does not have sufficient free space, it is determined whether the existing SLU can be moved to another control software 53, and if possible, the SLU is moved and then a new SLU is created by the provisionally selected control software 53 (steps NO in S705 to YES in S706 to S707 to S708 to S710), and if the SLU cannot be moved, another control software 53 is selected as the destination for creating the new SLU according to a specific rule, and a new SLU is created (steps NO in S705 to NO in S706 to S709 to S710).

[0193] As described above with reference to Fig. 24, the SLU creation process in the fifth embodiment is characterized in that, when selecting the control software 53 to be used for creating the SLU, the control software 53 operating in the node specified by the APP node ID is selected as the destination for creating the new SLU as much as possible. When the SLU 52 is created by the control software 53 operating in the node specified by the APP node ID, the storage OS VM 83 in which the SLU 52 exists and the user APP VM 82 that uses the SLU 52 exist in the same node, and therefore IO processing for the SLU 52 is performed via the network within the hypervisor. Therefore, the information processing system 8 according to the fifth embodiment can obtain high IO performance and reliability, similar to the information processing system 1 according to the second embodiment.

[0194] In addition, although detailed explanation will be omitted to avoid repetition, the information processing system 8 of the fifth embodiment has various programs in the VM 83 for the storage OS that perform operations similar to those of the storage node 3 in the first, third and fourth embodiments, thereby achieving the same effects as those of the first, third and fourth embodiments.

[0195] As a modified example of the information processing system 8 according to the fifth embodiment, a container configuration may be used instead of HCI. HCI is based on a hypervisor, but containers have a configuration in which an OS runs on a bare metal server, and a set of processes equivalent to a storage OS (storage OS container) and a user APP container run on the OS. Even with this container configuration, the present invention can be applied because the VM and storage are arranged on the same node in the same way as with HCI. [Explanation of symbols]

[0196] 1,8 Information Processing Systems 2 Host 3 Storage Node 4. Storage Service Network 5. Backend Network 6 Cluster 20,30 Internal Network 21,31 CPU 22,32 Memory 23,33 Storage device 24,34,35 Communication equipment 41 Virtual Machines (VMs) 51 ALU 52 SLU 53 Control Software 54 Cluster Control Unit 55 Configuration Management Department 56 Redundancy Group (Storage Controller Group) 60 Management information 61 Configuration Information 62 ALU Management Table 63,71,91 SLU Management Table 64 System Management Table 65 Control Software Management Table 66 Host Management Table 67 Path Management Table 72 System operation information management table 80 Hypervisor Nodes 81 Hypervisor Management Department 82 VM for user APP 83 VMs for Storage OS 84 User App IT Initiator LU Logical Unit TG Target

Claims

1. An information processing system including a plurality of storage nodes, each of which includes one or more processors, and a storage device, the plurality of storage nodes are connected via a network; The storage node includes: a first logical unit which is a logical volume; a second logical unit having a connection relationship with one or more of the first logical units and serving as a receiving port for an I / O request from a host device for data stored in the first logical unit having the connection relationship; a control unit that runs on the processor and processes data stored in the first logical unit; having a plurality of the control units implemented in different storage nodes are managed as a redundancy group, and one or a plurality of the first logical units are associated with the redundancy group; a pool based on a storage area of ​​the storage device is managed by the redundancy group, and the redundancy group creates the first logical unit using the pool that it manages; the plurality of control units constituting the redundancy group include an active control unit that processes the IO request to the first logical unit associated with the redundancy group, and a standby control unit that takes over the processing when a failure occurs in the active control unit; the second logical unit is disposed in a storage node in which the active control unit exists; When the processor receives an instruction to create the first logical unit, it designates the active control unit in accordance with a predetermined rule to create the first logical unit, and sets the connection relationship between the second logical unit, which was placed before the instruction to create the first logical unit in the storage node in which the designated active control unit exists, and the first logical unit created in response to the instruction to create the first logical unit. An information processing system comprising:

2. When arranging the redundancy group in the storage node, the processor a second logical unit creation process for creating the second logical unit in the same storage node as the active control unit of the redundancy group; a path setting process for setting a path serving as an I / O path between the second logical unit created in the second logical unit creation process and the host device; a first logical unit creation process for creating the first logical unit in the same storage node as the active control unit; a bind process for establishing a connection relationship between the first logical unit created in the first logical unit creation process and the second logical unit in the same storage node; 2. The information processing system according to claim 1, wherein the information processing system executes the steps of:

3. In the path setting process, the processor setting a plurality of paths between the host device and the second logical unit, the paths passing through any of ports of a plurality of storage nodes; the ports through which the multiple paths pass are present in multiple storage nodes, The storage nodes in which the ports through which the multiple paths pass include the storage node in which the active control unit is located and the storage node in which the standby control unit is located.

3. The information processing system according to claim 2.

4. In the path setting process, the processor Among the plurality of paths that have been set, a path that passes through a port of the storage node in which the active control unit that controls the second logical unit exists is set as a priority path to be used preferentially; Among the plurality of paths that have been set, a path that passes through a storage node in which the standby control unit that takes over the processing of the active control unit exists when a configuration change occurs due to a failover is set as a redundant path that is changed to the preferred path when the failover occurs.

4. The information processing system according to claim 3.

5. In the first logical unit creation process, the processor An active control unit of the redundancy group in which the first logical unit is to be created is selected based on the capacity of the first logical unit to be created and the free capacity of the redundancy group.

3. The information processing system according to claim 2.

6. In the binding process, the processor The connection relationship is established between the first logical unit created in the first logical unit creation process and the second logical unit controlled by the same active control unit as the first logical unit.

3. The information processing system according to claim 2.

7. In the binding process, the processor When there are a plurality of second logical units controlled by the same active control unit as the first logical unit created in the first logical unit creation process, the connection relationship is set to the second logical unit having the least number of settings of the connection relationship with the first logical unit.

7. The information processing system according to claim 6.

8. In the first logical unit creation process, the processor When an application of the host device that accesses the new first logical unit to be created is accessing an existing first logical unit, the new first logical unit is created in association with the active control unit that controls the existing first logical unit.

3. The information processing system according to claim 2.

9. In the first logical unit creation process, the processor When the active control unit that controls the existing first logical unit does not have enough capacity to create the new first logical unit, another first logical unit in the storage node where the active control unit exists is moved to another storage node, and the new first logical unit is created in association with the active control unit.

9. The information processing system according to claim 8.

10. the processor is capable of executing the first logical unit creation process in response to an instruction to create the first logical unit; In the first logical unit creation process, the processor If there is no active control unit in the plurality of storage nodes in which the number of controls of the first logical unit is equal to or less than a predetermined number, an active control unit that has free capacity required for the new first logical unit specified in the creation instruction and has the fewest number of controls of the second logical unit is selected, the second logical unit creation process is executed to create the second logical unit, and the new first logical unit corresponding to the created second logical unit is created.

3. The information processing system according to claim 2.

11. the processor is capable of executing the bind process in response to a connection relationship setting instruction in which the first logical unit to be connected is designated; In the binding process, the processor When the number of the connection relationships set in the second logical unit controlled by the same active control unit as the first logical unit to be connected exceeds a predetermined number, the second logical unit is created, and the connection relationship with the first logical unit to be connected is moved to the created second logical unit.

3. The information processing system according to claim 2.

12. When the number of storage nodes is increased to N, the processor execute the second logical unit creation process so that the number of the second logical units arranged in each storage node is N-1, and execute the path setting process after setting a connection relationship between the second logical units created in the second logical unit creation process and the first logical units in the same storage node; When any of the N storage nodes is removed, the first logical unit in the storage node can be distributed and arranged in N-1 storage nodes other than the storage node.

3. The information processing system according to claim 2.

13. The processor, classifying the plurality of first logical units into a low-load logical unit group and a high-load logical unit group, and setting the connection relationship between each of the classified logical unit groups and different second logical units; a node load adjustment process is executed to move the low-load logical unit group and the second logical unit with which the connection relationship is established with the logical unit group to a storage node with a high load among the plurality of storage nodes, and to move the high-load logical unit group and the second logical unit with which the connection relationship is established with the logical unit group to a storage node with a low load among the plurality of storage nodes.

3. The information processing system according to claim 2.

14. An information processing system configured by connecting a plurality of computer nodes via a network, each of which includes one or a plurality of processors and a storage device, Each of the computer nodes is implemented with a storage virtual machine that realizes a function equivalent to one of the storage nodes, and a host device virtual machine that has an application that accesses any one of the first logical units and operates, and realizes a function equivalent to one of the host devices, as virtual machines that operate on a hypervisor; In the first logical unit creation process, the processor When the application that accesses the new first logical unit to be created is accessing an existing first logical unit, the new first logical unit is created in association with the active control unit that controls the existing first logical unit.

3. The information processing system according to claim 2.

15. A configuration management method for an information processing system including a plurality of storage nodes, each of which includes one or a plurality of processors, and a storage device, comprising: the plurality of storage nodes are connected via a network; The storage node includes: a first logical unit which is a logical volume; a second logical unit having a connection relationship with one or more of the first logical units and serving as a receiving port for an I / O request from a host device for data stored in the first logical unit having the connection relationship; a control unit that runs on the processor and processes data stored in the first logical unit; having a plurality of the control units implemented in different storage nodes are managed as a redundancy group, and one or a plurality of the first logical units are associated with the redundancy group; a pool based on a storage area of ​​the storage device is managed by the redundancy group, and the redundancy group creates the first logical unit using the pool that it manages; the plurality of control units constituting the redundancy group include an active control unit that processes the IO request to the first logical unit associated with the redundancy group, and a standby control unit that takes over the processing when a failure occurs in the active control unit; the second logical unit is disposed in a storage node in which the active control unit exists; When the processor receives an instruction to create the first logical unit, it designates the active control unit in accordance with a predetermined rule to create the first logical unit, and sets the connection relationship between the second logical unit, which was placed before the instruction to create the first logical unit in the storage node in which the designated active control unit exists, and the first logical unit created in response to the instruction to create the first logical unit. A configuration management method comprising:

Citation Information

Patent Citations

  • Storage system and method for controlling the same

    JP2019101702A

  • Information processing system and path management method

    JP2019185328A

  • Method to determine disposition for VM / container and volume in HCI environment, and storage system

    JP2020052730A

  • Storage system, path control method, and path control program

    JP2021056567A

  • Storage system, storage device and access control method

    WO2016103416A1