Managing method for storing subsystem

CROSS-REFERENCES TO RELATED APPLICATIONS

0001This application relates to and claims priority from Japanese Patent Application No. 2004-124385, filed on April 20, 2004, the entire disclosure of which is incorporated herein by reference.

BACKGROUND OF THE INVENTION

1. Field of the Invention

0002The present invention relates to a storage subsystem, which stores data of a computer system, and an integrated system of the storage subsystem and the computer system. In particular, the present invention relates to a technique for load distribution and failure avoidance among interface ports in the case inwhich the storage subsystem includes plural interface ports.

2. Description of the Related Art

0003As a system that manages a large amount of electronic data, there is a storage subsystem represented by a disk array apparatus. In the storage subsystem, it is necessary to perform failure avoidance and load distribution processing in order to utilize electronic data efficiently.

0004Conventionally, in information processing system having a host computer and a storage subsystem, there has been known a device program that performs load distribution and backward processing between the host computer and the storage subsystem (e.g., Power Path produced by EMC Corporation).

SUMMARY OF THE INVENTION

0005Incidentally, a large storage subsystem has a problem in that an administrator is required to set all combinations of logic devices and ports of the storage subsystem, and management is complicated.

0006However, in the above-mentioned device program, integrated management in the storage subsystem is not specifically taken into account.

0007In addition, in the device program, in the case in which load distribution processing and failure avoidance processing are performed, it is necessary to install a device program, which performs the load distribution processing and the failure avoidance processing, on a host computer connected to the storage subsystem. However, the device program is designed so as to perform the load distribution processing and the failure avoidance processing in an upper layer of a layer, in which a general-purpose protocol is used, in accordance with a unique protocol.

0008Therefore, in the device program, a problem of compatibility between the host computer and the storage subsystem occurs. More specifically, in the case in which load distribution and failure avoidance processing for a storage subsystem, to which plural disk array apparatuses of different manufacturers are connected, are performed, it is necessary to install a device program for the storage subsystem for each of the manufacturers on a host computer.

0009In order to solve the above-mentioned problem, an aspect of the present invention is applied to a management computer that allocates ports to an information processing system including at least one host computer and at least one storage subsystem that is connected to the host computer via a network.

0010Here, the storage subsystem has plural logic devices and plural ports and is connected to the network via the plural ports. The management computer executes: the step of receiving input of data indicating one or more storage domains that are management units for the storage subsystem; the step of grouping the plural ports, the plural logic devices, and the host computers in the one or more storage domains; the step of, in the case in which a discovery request for the logic devices is received from the host computer, specifying a storage domain to which both the host computer, which has sent the discovery request, and the logic devices, for which the discovery is requested, belong out of the storage domains in which the plural ports, the plural logic devices, and the host computers are grouped; the step of selecting ports, which belong to the specified storage domain, and allocating the selected ports to the logic devices; and the step of sending information on the allocated ports to the host computer.

BRIEF DESCRIPTION OF THE DRAWINGS

0011In the accompanying drawings: <ul id="ul0001" list-style="none" compact="compact"><li>Fig. 1 is a block diagram showing an example of a system configuration of an information processing system according to a first embodiment of the present invention;</li><li>Fig. 2 is a block diagram showing an example of a system configuration of a storage subsystem 102 according to the first embodiment;</li><li>Fig. 3 is a diagram illustrating a storage domain that is a management unit for the storage subsystem;</li><li>Fig. 4 is a diagram schematically showing a data structure of a storage domain management table 113;</li><li>Fig. 5 is a diagram schematically showing a data structure of a port management table 112;</li><li>Fig. 6 is a diagram schematically showing a data structure of a device management table 111;</li><li>Fig. 7 is a diagram schematically showing a data structure of a host management table 116;</li><li>Fig. 8 is a diagram schematically showing a data structure of a name management table 115;</li><li>Fig. 9 is a diagram for explaining a flow of initial setting processing that is performed by the information processing system;</li><li>Fig. 10 is a diagram for explaining a flow of logic device creation processing that is performed by the information processing system;</li><li>Fig. 11 is a diagram for explaining a flow of port allocation processing;</li><li>Fig. 12 is a diagram for explaining a flow of configuration information update processing that is performed by the information processing system;</li><li>Fig. 13 is a diagram for explaining a flow of discovery processing that is performed by a management computer 102;</li><li>Fig. 14 is a diagram for explaining a flow of port allocation processing that is performed by the management computer 102;</li><li>Fig. 15 is a block diagram showing an example of a system configuration of an information processing system according to a second embodiment of the present invention;</li><li>Fig. 16 is a diagram for explaining a SAN domain;</li><li>Fig. 17 is a diagram schematically showing a data structure of a SAN domain management table 1503;</li><li>Fig. 18 is a diagram schematically showing a data structure of a port management table 1504;</li><li>Fig. 19 is a diagram schematically showing a data structure of a host management table 1505;</li><li>Fig. 20 is a diagram for explaining a flow of initial setting processing that is performed by the information processing system according to the second embodiment;</li><li>Fig. 21 is a diagram for explaining a flow of port allocation processing that is performed by a management computer according to the second embodiment;</li><li>Fig. 22 is a diagram for explaining port allocation processing that is performed by a management computer 103 according to a third embodiment of the present invention;</li><li>Fig. 23 is a diagram for explaining a flow of port allocation processing that is performed by an information processing system according to the third embodiment;</li><li>Fig. 24 is a diagram for explaining a flow of configuration update processing that is performed by the information processing system according to the third embodiment;</li><li>Fig. 25 is a block diagram showing an example of a system configuration of an information processing system according to a fourth embodiment of the present invention; and</li><li>Fig. 26 is a diagram for explaining a flow of port allocation processing that is performed by the information processing system according to the fourth embodiment.</li></ul>

DESCRIPTION OF THE PREFERRED EMBODIMENTS

0012First, a first embodiment of the present invention will be explained.

0013Fig. 1 is a block diagram showing a system configuration of an information processing system according to the first embodiment. Fig. 2 is a system diagram of a storage subsystem 102 according to the first embodiment.

0014As shown in Fig. 1, the information processing system includes a host computer 101 serving as a host apparatus, a management computer 103, and the storage subsystem 102. The host computer 102, the management computer 103, and the storage subsystem 102 are connected to a storage area network (SAN) 104, respectively. The host computer 101, the management computer 103, and the storage subsystem 102 exchange information via the storage area network (SAN) 104. Note that the host computer 101 and the storage subsystem 102 may be provided in a single form or a plural form.

0015The host computer 101 inputs and outputs data to the storage subsystem 102 and the management computer 103, respectively, via the SAN 104. A workstation, a microcomputer, or a mainframe computer can be used as the host computer 101. Application programs and data base systems are active in the host computer 101.

0016The management computer 103 manages the storage subsystem 102, the SAN 104, and the host computer 101. More specifically, the management computer 103 includes a processor (CPU: Central Processing Unit) 10, a memory 20, and an FCI/F 11. In the memory 20, a name management program 114 , a storage management program 110a, a device management table 111a, a port management table 112a, a storage domain management table 113a, a name management table 115, and a host management table 116 are stored.

0017The FCI/F 11 controls data transmission and reception that is performed between the management computer 103 and the host computer 101 via the SAN 104. In addition, the FCI/F 11 controls data transmission and reception that is performed between the management computer 103 and the storage subsystem 102 via the SAN 104.

0018The CPU 10 executes the storage management program 110a to manage the storage subsystem 102, the SAN 104, and the host computer 101. In addition, the CPU 10 executes the name management program 114 to provide a name service to the host computer 101 and the storage subsystem 102 that are connected to the information processing system.

0019The storage management program 110a is a program that manages the storage subsystem 102 in the system. Information on the storage subsystem 102 in the system is stored in the device management table 111a, the port management table 112a, and the storage domain management table 113a.

0020The CPU 10 executes the storage management program 110a to thereby periodically acquire various kinds of information on the storage subsystem 102 from a storage management program 110b to be described later that operates in the entire storage subsystem 102 in the system. The various kinds of information to be acquired from the storage subsystem 102 may be sent to the management computer 103 every time the storage management program 110b, which operates in the storage subsystem 102, updates the various kinds of information.

0021The CPU 10 also executes the storage management program 110a to manage the host management table 116. In the host management table 116, the host computer 101 in the system and storage domains are stored in association with each other.

0022The name management program 114 is a program for providing a name service to the host computer 101 and the storage subsystem 102 using existing protocols such as iSNS (Internet Storage Name Service). The CPU 10 executes the name management program 114 to communicate with the host computer 101 and the storage subsystem 102 and manage the name management table 115. In the name management table 115, protocol information concerning the storage subsystem 102, logic devices 105 included in the storage subsystem 102, the host computer 101, and the like is stored. For example, in the case in which iSNS is used as the protocol for the name management program 114, in the name management table 115, information such as "iSCSI NAME", "IP address", and "Discovery Domain" is stored as regards iSCSI (Internet SCSI).

0023In this embodiment, data concerning the storage subsystem 102 to be stored in the name management table 115 is updated as described below. The CPU 10 executes the storage management program 110a to acquire the data on the storage subsystem 102 from the storage management program 110b operating in the storage subsystem 102 (a CPU 202 that executes the storage management program 110b) and store the acquired data in a work area of the memory 20. Thereafter, the CPU 10 executes the name management program 114 to read out the data concerning the storage subsystem 102 stored in the work area and update the data in the name storage table 115 using the read-out data.

0024The storage subsystem 102 is a storage subsystem that includes a control unit (a controller 201 shown in Fig. 2), which controls exchange of information with the host computer 101, and a disk device (a disk device 210 shown in Fig. 2), which stores information. This embodiment will be explained with an example in which a disk array apparatus is applied as the storage subsystem 102 as an example. However, the present invention is not specifically limited to this. A storage subsystem other than the disk array apparatus may be applied to the storage subsystem 102.

0025The storage subsystem 102 has one or more ports 204 and is connected to the host computer 101 through the SAN 104. The ports 204 are interfaces for a storage subsystem such as a fiber channel, iSCSI (Internet SCSI), FICON, or ESCON. Note that, in this embodiment, a structure of the SAN 104 and a type of the Interface 104 are not specifically limited. Note that this embodiment will be explained with a case in which the iSCSI is applied to the ports 204 and the IP-SAN is applied to the SAN as an example.

0026The storage subsystem 102 has one or more logic devices 105. The host computer 101 is capable of accessing the logic devices 105, for example, reading data from and writing data in the logic devices 105 via the ports 204. In this case, the host computer 101 uses a protocol for a storage subsystem provided by the respective ports 204. iSCSI, FCP (Fibre Channel Protocol for SCSI), and FC-SB (Single Byte Protocol) are used in the IP-SAN, the fiber channel, and the FICON, respectively. Note that a type of the protocol used in this embodiment is not specifically limited.

0027Next, a structure of the storage subsystem 102 will be described in detail using Fig. 2.

0028As shown in Fig. 2, the storage subsystem 102 includes a controller 201 and one or more disk devices 210. The controller 201 includes a host adapter 203, a memory 205, disk adapters 207, and a processor (CPU: Central Processing Unit) 202. The numbers of the respective devices are preferably multiplexed from the viewpoint of performance and reliability.

0029The host adapter 203 controls a protocol concerning the ports 204 such as the iSCSI.

0030Programs and data concerning storage domain control are stored in the memory 205. More specifically, in the memory 205, the storage management program 110b, a device management table 111b, a port management table 112b, and a storage domain management table 113b are stored.

0031In addition, in the memory 205, other than these programs and data, a control program and control information necessary for controlling the storage subsystem 102, and cache data 110 concerning input/output processing for the host computer 101 are stored. It is preferable to multiplex the memory 205 or multiplex a power supply to the memory 205 in order to realize high reliability.

0032The disk adapters 207 perform protocol processing concerning a disk interface 209 such as a fiber channel in the same manner as the host adapter 203.

0033The disk devices 210 receive a command for reading, writing, or the like from the controller 201 via the disk interface 209 and perform processing defined by the command. It is preferable to duplex the disk interface 209 in order to realize high reliability.

0034In the storage subsystem 102, the plural disk devices 210 are combined to form a redundant structure, in which logic devices are created, and the logic devices 105 (see Fig. 1) are created by one or more logic devices.

0035The CPU 202 executes processing concerning the storage subsystem 102. More specifically, the CPU 202 is connected to the host adapter 203 and the disk adapters 207 in the controller 201 via an internal bus 208. The CPU 202 controls the host adapter 203 and the disk adapters 207. In addition, the CPU 202 is connected to the memory 205 via the internal bus 208, and loads and executes the programs and the data (the storage management program 110b, the device management table 111b, the port management table 112b, and the storage domain management table 113b) stored in the memory 205.

0036The storage management program 110b is a program for managing information on other resources such as the ports 204 and the logic devices 105 concerning the storage subsystem 102 using the respective tables (the device management table 111b, the port management table 112b, and the storage domain management table 113b).

0037The device management table 111b is a table for managing association of attributes of the devices 105 and the storage domains. The port management table 112b is a table for managing association of the ports 204 of the storage subsystem 102 and the storage domains. The storage domain management table 113b is a table for managing attributes of the storage domains.

0038Note that the storage management program 110b explained above is installed in a storage medium, which is provided in each storage subsystem 102, from other devices using a portable storage medium such as a compact disk or a magneto-optical disk or via a network.

0039Next, the storage domain, which is a management unit for the information processing system according to the first embodiment, will be explained using Fig. 3.

0040Fig. 3 is a diagram for explaining the storage domain that is a management unit for the storage subsystem according to this embodiment.

0041In storage domains 301, a group of the logic devices 105 and a group of the ports 204 of the storage subsystem 102, and the host computers 101 are classified. The host computers 101 are classified according to attributes such as departments in which the host computers 101 are used, performance, physical connection, and the like. Note that the storage domains 301 may be managed hierarchically using attributes such as departments, application programs, users, and the like.

0042The group of the logic devices 105, the group of the ports 204, and the host computers 101, which are classified in the same storage domains 301, can be combined arbitrarily. That is, the host computers 101 can access the logic devices 105 belonging to the same storage domain through arbitrary ports belonging to the same storage domain.

0043In the example shown in the figure, the group of the logic devices 105, the group of the ports 204, and the host computers 101 in which the information processing systemhas are classified in two storage domains (301a and 301b). More specifically, logic devices 105a and 105b, ports 204a and 204b, host computers 101a and 101b belong to a storage domain 301a of a storage subsystem A. A logic device 105c, ports 204c and 204d, and a host computer 101c belong to the storage domain 301b.

0044Since the host computer 101a belongs to the storage domain 301a, the host computer 101a is capable of accessing the logic devices 105a and 105b. In accessing the logic devices 105a and 105b, the host computer 101a can use the port 204a and 204b.

0045Since the host computer 101c belongs to the storage domain 301b, the host computer 101c is capable of accessing the logic device 105c. In accessing the logic device 105c, the host computer 101c can use the ports 204c and 204d.

0046Specific criteria for classifying the group of the logic devices 105, the group of the ports 204, and the host computer 101 in the storage domains 304 are not limited in this embodiment. For example, as the criteria for classification, security concerning propriety of access and the like, performance and reliability concerning throughput and delay at the port 204 and the SAN 104,and the like may be used.

0047In the case in which the classification is performed using the criteria concerning security, association with an existing user management program or the like is possible. For example, as the existing user management program, association with an Active Directory (registered trademark) server installed in Windows (registered trademark), which is an OS provided by Microsoft Corporation, is possible. In addition, for example, association with an LDAP (Lightweight Directory Access Protocol) server and an NIS (Network Information Service) server is possible in a UNIX environment.

0048For example, in the iSCSI, iSNS or SLP (Service Location Protocol) can also be used. In an iSNS server, association with the LDAP or the like is possible, and a management classification such as departments can usually be succeeded.

0049Concerning the criteria concerning performance, it is advisable to take into account the following description. For example, in a company or the like, it may be required to secure performance for applications such as a database and special users such as an administrator even in an identical department. Thus, as in the case of the security management, it is advisable to provide classification concerning performance. For example, applications and users with low priorities and applications and users with high priorities are classified in separate storage domains 301. The storage domains 301 are set such that the ports 204 are allocated to specific applications and a port is shared by the other applications.

0050Topology of a network may affect performance. Forexample, path from an arbitrary host to the respective ports 204 in the identical storage domain 301 may be different. In this case, performance (throughput, delay) is different. Therefore, in the case in which an IP network is used, a sub-network of an IP address may be analyzed to perform classification. Alternatively, a command for tracing a network path (e.g., a traceroute command in the Unix environment) may be used to perform classification on the basis of paths from ports in the identical storage domain 301.

0051In this way, in this embodiment, the group of logic devices 105, the group of the ports 204, and the host computers 101 are classified in the storage domains 301a such that the logic devices and the ports belonging to the same storage domain can be combined arbitrarily. Therefore, it becomes unnecessary to perform setting for associating logic devices and ports in a one-to-one correspondence as in the conventional storage subsystem, and complexity of management of the storage subsystem can be eliminated.

0052In addition, in this embodiment, the group of the logic devices 105, the group of the ports 204, and the host computers 101 can be classified according to attributes such as departments, application programs, and users. This makes it possible to access a logic device corresponding to an attribute of a user efficiently.

0053Next, data structures of the respective tables (the storage domain management table 113, the port management table 112, the device management table 111, the host management table 116, and the name management table 115) included in the information processing system according to this embodiment will be described using Figs. 4 to 8.

0054Fig. 4 is a diagram schematically showing a data structure of the storage domain management table 113 according to this embodiment.

0055The storage domain management table 113 is a table for holding attributes of the storage domains 301.

0056The storage domain management table 113 includes an entry 401 in which a "storage domain ID", which is identifier of a storage domain, is stored, an entry 402 in which a "storage domain name" indicating a name for management of the storage domain is stored, and an entry 403 in which "attribute data" indicating an attribute of the storage domain is stored. The "storage domain names" and the "attribute data" are stored in the entries 402 and 403 in association with the "storage domain IDs" stored in the entry 401.

0057In the example shown in the figure, data with the "storage domain ID" of "sa", the "storage domain name" of "Account", and the "attribute data" of "accounting" is stored. In addition, data with the "storage domain ID" of "sb", the "storage domain name" of "Develop", and the "attribute data" of "development" is stored.

0058The CPU 10 of the management computer 103 executes the storage management program 110a to, for example, display a storage domain setting screen on a display device (not shown). The CPU 10 of the management computer 103 executes the storage management program 110a to receive definitions of the storage domains (setting for "storage domain IDs", "storage domain names", and "attribute data") from an administrator via an input device (not shown). The CPU 10 of the management computer 103 executes the storage management program 110a to store the received definitions in the storage domain management table 113.

0059Note that, in the above description, the CPU 10 of the management computer 103 executes the storage management program 110a to receive the definitions of the storage domains from the administrator and store the received definitions in the storage domain management table 113. However, the CPU 202 of the storage subsystem 102 can also execute the storage management program 110b to perform the same processing as above.

0060In addition, the administrator may input the "attribute data" and the "storage domain names" by operating the storage management program 110a stored in the management computer 103. However, efficiency is improved if the "attribute data" and the "storage domain names" are set in association with the management program such as Active Directory, LDAP, or NIS.

0061Fig. 5 is a diagram schematically showing a data structure of the port management table 112 according to this embodiment.

0062The port management table 112 is a table for associating the ports 204, the storage subsystem 102, and the storage domains 301.

0063The port management table 112 includes an entry 501 in which an "apparatus ID" serving as an apparatus identifier for the storage subsystem 102 is stored, an entry 502 in which a "port ID" serving as an identifier for the port 204 is stored, an entry 503 in which an "address" on the SAN 104 of each of the ports 204 is stored, an entry 504 in which a "utilization rate" of the port 204 is stored, an entry 505 in which "state data" indicating a state of the port is stored, and an entry 506 in which a "storage domain ID" of the storage domain 301 to which the port 204 belongs is stored.

0064The respective data are stored in the entries 502 to 506 in association with the "apparatus ID" stored in the entry 501. The respective data are stored in the entries 503 to 506 in association with the entry 502.

0065Note that, in this embodiment, one storage domain 301 is allocated to one port. However, it is possible to allocate plural storage domains 301 to any port. In the case in which the plural storage domains are allocated to one port, plural storage domain IDs , which are stored in the entry 506, are set. For example, in the case of the IP network, it is advisable to use IP addresses and port numbers for the addresses stored in the entry 503.

0066In the example shown in the figure, "A1" , "90%", "overload" , and "sa" are stored in the entries 503 to 506 corresponding to the entry 502 in which "A" is stored, respectively. This indicates that a port A included in the storage subsystem 102 with the apparatus ID of "storage A" has an "address" on the SAN 104 of "A1" and belongs to the storage domain 301 having the storage ID of "sa". In addition, this indicates that the port A has a port utilization rate of "90%" and is currently in an overload state.

0067The CPU 10 of the management computer 103 (or the CPU 202 of the storage subsystem 102) executes the storage management program 110a (or the storage management program 110b) to perform management of the port management table 112. The CPU 10 of the management computer 103 executes the storage management program 110a to associate the apparatus, the ports, and the storage domains 301 on the port management table 112.

0068The CPU 10 of the management computer 103 executes the storage management program 110a to periodically acquire a load (port utilization rate) and a configuration state of the ports 204 of the storage subsystem 102 and judge a current "state" on the basis of the acquired "port utilization rate". The CPU 10 of the management computer 103 (or the CPU 202 of the storage subsystem 102) stores the calculated "port utilization rate" and the judged "state" in corresponding entries in the port management table 112.

0069Note that, although a form of the "state" to be judged is not limited, for example, states of "overload", "normal", "offline" , and "blocked" may be indicated as the form of the "state" to be judged. Then, the CPU 10 of the management computer 103 acquires a configuration of the storage subsystem 102 periodically or every time the configuration is changed to thereby set states of "normal", "offline", and "blocked". Although the judgment on overload is different depending upon a system, in general, it is advisable to set about 80% to 90% as a threshold value.

0070The CPU 10 of the management computer 103 performs port allocation processing to create an available port list on the basis of information in the port management table 112 (port allocation processing will be described later).

0071Fig. 6 is a diagram schematically showing a data structure of the device management table 111 according to this embodiment.

0072The device management table 111 includes an entry 601 in which an "apparatus ID" serving as an apparatus identifier for the storage subsystem 102 is stored, an entry 602 in which a "device ID" specifying the logic device 105 in the storage subsystem 102 is stored, and an entry 607 in which a "storage domain ID" serving as an identifier for the storage domain 301 corresponding to each of the logic device is stored.

0073The device management table 111 includes an entry 603 in which a "protocol" for the logic device 105 is stored, an entry 604 in which a "size" indicating a capacity of the logic device 105 is stored, an entry 605 in which "authentication information" indication information for authentication for the logic devices 105 is stored, and an entry 606 in which a "state" indicating whether the logic device 105 is "used" or "not used" is stored such that general attributes of the respective logic devices 105 can also be managed.

0074Note that, in this embodiment, the protocols to be stored in the entry 603 are not specifically limited. For example, "iSCSI", "FCP", and "SCSI" can be used for the protocols. As the authentication information in the entry 605, authentication information on the respective devices 105 in the "iSCSI", for example, key values, passwords, and the like are stored.

0075The device management table 111 is managed by the CPU 10 of the management computer 103 that executes the storage management program 110a (or the CPU 202 of the storage subsystem 102 that executes the storage management program 110b). More specifically, the CPU 10 of the management computer 103 (or the CPU 202 of the storage subsystem 102) executes the storage management program 110a (or the storage management program 110b) to store data corresponding to the respective entries 601 to 607. The CPU 10 of the management computer 103 (or the CPU 202 of the storage subsystem 102) monitors states of use of the logic devices periodically to update the data in the entry 606.

0076Fig. 7 is a diagram schematically showing a data structure of the host management table 116 in this embodiment.

0077As shown in the figure, the host management table 116 includes an entry 701 in which an "apparatus ID" serving as an apparatus identifier for the host computer 101 is stored, an entry 702 in which a "port ID" identifying a port of the host computer 101 is stored, an entry 703 in which an "address" allocated to each of the ports is stored, an entry 704 in which a "protocol" supported by each of the ports is stored, an entry 705 in which a "state information" indicating a state of the port of the host computer 101 is stored, and an entry 706 in which a "storage domain ID" serving as an identifier for the storage domain 301 corresponding to the port of the host computer 101 is stored. Note that, in the "state information" indicating the state of the port, data indicating whether the state of the port is "normal" or "abnormal" is stored.

0078The host management table 116 is managed by the CPU 10 of the management computer 103 that executes the storage management program 110a. The CPU 10 of management computer 103 executes the storage management program 110a to store corresponding data in the respective entries 701 to 706. The CPU 10 periodically acquires the data indicating states of the ports 204 of the respective hosts from the storage subsystem 102 and stores the data indicating the "state information" in the entry 705.

0079Fig. 8 is a diagram schematically showing a data structure of the name management table 115 according to this embodiment. Name management information in the SAN 104 is stored in this name management table 115.

0080The name management table 115 is managed by the CPU 10 of the management computer 103 that executes a name management program. The CPU 10 of the management computer 103 executes the name management program to provide a name service to the host computer 101 and the storage subsystem 102 utilizing data in the name management table 115. This embodiment will be described with iSNS, which is a representative name server of the iSCSI, as an example of the name management program 114.

0081As shown in the figure, the name management table 115 includes an entry 801 in which an "EID (Entity ID)" serving as an identifier for iSNS is stored, an entry 802 in which an "iSCSI Name" serving as an identifier for an iSCSI node is stored, an entry 803 in which a "DD (Discovery Domain)" defining a name service range in the iSNS is stored, an entry 804 in which a "Node Type" indicating a type such as Initiator/Target of the iSCSI nodes is stored, and entries 805 and 806 in which a "portal Group" and a "Portal Address: Port" serving as information on the port 204 are stored, respectively.

0082Respective iSCSI devices (e.g., the host computers 101) request the CPU 10 of the management computer 103, which executes the name management program 114, to register the iSCSI nodes when an administrator instructs to do so or starts the iSCSI devices. The information requested to be registered include the "DD (Discovery Domain) " , the "Node Type", the "Portal Group" , and the "Portal Address: Port".

0083In an information processing system including a normal storage subsystem, the host computer 101 and the storage subsystem 102 request the CPU 10 of the management computer 103, which executes the name management program 114, to register the iSCSI node individually. However, in this embodiment, concerning data on the respective logic devices 105, the CPU 202 of the storage subsystem 102 executes the storage management program 110b to send configuration information to the CPU 10 of the management computer 103. The CPU 10 of the management computer 103 executes the storage management program 110a to receive the sent configuration information and store the received configuration information in a work area of the memory 20. Thereafter, the CPU 10 of the management computer 103 executes the name management program 114 to read out the stored configuration information and register the configuration information in the name management table 115.

0084The CPU 10 executes the name management program 114 to perform processing for giving an EID (Entity ID), which is an identifier for an iSNS, to an iSCSI node to register the iSCSI node in the entry 801. The respective iSCSI devices are capable of requesting the CPU 10 of the management computer 103 to register data in the name management table 115, change the data, and delete the data from the name management table 115 using the EID. In this embodiment, the management is facilitated if an arbitrary "DD" is associated with one of the storage domains 301.

0085In Figs. 4, 5 and 6, general information is shown. In the following explanation, correspondence in a case of an iSNS server will be described. More specifically, the "Portal Address: Port" stored in the entry 806 corresponds to the address in the entry 503 shown in Fig. 5. The "iSCSI Name" in the entry 802 corresponds to the "device ID" in the entry 502 shown in Fig. 5.

0086Fig. 9 is a diagram for explaining a flow of initial setting processing that is performed by the information processing system according to this embodiment.

0087The information processing system according to this embodiment performs initial setting processing when the system is installed.

0088First, the CPU 10 of the management computer 103 of the information processing system executes the storage management program 110a to receive setting for the storage domains 301 from an administrator (S901). More specifically, the CPU 10 of the management computer 103 executes the storage management program 110a to display a setting screen on a display device (not shown) and receive the setting for the storage domains 301 from the administrator via an input device (not shown). The CPU 10 executes the storage management program 110a to add or update the received information on the storage domains in the storage domain management table 113 and the port management table 112. Note that the CPU 10 of the management computer 103 transfers information on the storage domains and the like, which is inputted from the administrator, to the storage subsystem 102 via the network. The controller 203 of the storage subsystem 102 adds or updates the information in the storage domain management table 113b, which is included in the controller 201, on the basis of the received information.

0089Note that, in the above explanation, the administrator sets the storage domains 301 manually. However, the present invention is not limited to this. Other various forms are possible. For example, the administrator inputs the storage domains 301 with reference to management information of other management programs, or a storage management program directly refers to the management information of the other management programs to propose the storage domains 301 to the administrator, and the administrator corrects the storage domains 301.

0090The CPU 10 of the management computer 103 creates the logic devices 105 after setting the storage domains 301 (S902) . Device creation processing in this embodiment is substantially the same as the device creation processing of the storage subsystem 102 according to the conventional technique. The device creation processing in this embodiment is different from the device creation processing in the conventional technique only in that logic devices are allocated to the storage domains 301. Note that a specific procedure for the device creation processing will be described later. This device creation processing is carried out every time a device is created.

0091Fig. 10 is a diagram for explaining a flow of logic device creation processing that is performed by the information processing system according to this embodiment. Note that, in the following explanation, a case in which the logic device creation processing is executed by the management computer 103 will be described. However, the same processing can be performed by the storage subsystem 102. In addition, the CPU 10 of the management computer 103 executes the storagemanagement program 110a, whereby the following processing shown in the figure is realized. However, for simplification of the explanation, it is assumed that the storage management program 110a is assumed to be a subject of operation.

0092First, an administrator inputs a logic device creation request to the management computer 103 via an input device. The storage management program 110a receives the request for device creation from the administrator (S1001). Note that, more specifically, the request for device creation means information shown on the device management table 111 (a size and authentication information of the logic device 105 that is desired to be created, the storage domain 310 that is desired to be allocated to the logic device 105, etc.).

0093The storage management program 110a analyzes the information received in S1001 to judge whether it is necessary or unnecessary to create the logic device 105 (S1002). Then, if it is judged that it is necessary to create the logic device 105, the storage management program 110a proceeds to S1003 and S1004 . On the other hand, if it is judged that it is unnecessary to create the logic device 105, the storage management program 110a proceeds to S1005. The case in which it is unnecessary to create the logic device 105 means a case in which the logic device 105 matching attributes requested in S1001 is already present, and it is unnecessary to create the logic device 105 anew.

0094In S1003 and S1004, the storage management program 110a carries out processing for creating the logic device 105. In the case in which there is the unallocated logic devices 105, which match the attributes requested in S1001, or an unused area in the storage subsystem 102 sufficiently, the storage management program 110a creates the logic device 105. More specifically, the storage management program 110a instructs the storage subsystem 102 to create the requested logic device 105. The storage subsystem 102, which has received the instruction, creates the logic device 105 based upon requested contents and informs the management computer 103 of a result of the creation. If the logic device 105 has been successfully created, the storage management program 110a proceeds to processing in S1005. On the other hand, if it is impossible to create the logic device 105 matching the requested attributes, the storage management program 110a performs error report (S1013) to end the processing. Note that specific means for the error report performed by the storage management program 110a is not specifically limited. For example, an error screen is displayed on a display device.

0095In S1005, the storage management program 110a performs processing for the storage domains 301. More specifically, the storage management program 110a judges whether or not an instruction for the storage domain 301, to which the logic device 105 is allocated (allocation storage domain 301), is included in the request received in S1001. If the instruction for the storage domain 301 has not been received, the storage management program 110a sets the default storage domain 301 as the allocation storage domain 301 and proceeds to S1007. On the other hand, if the instruction about the storage domain 301 has been received, the storage management program 110a proceeds to S1006.

0096In S1006, the storage management program 110a sets the instructed storage domain as the allocation storage domain 301 and proceeds to S1007.

0097In S1007, the storage management program 110a judges whether or not the allocation storage domain 301 is present on the basis of the storage domain management table 113. If the allocation storage domain 301 is present, the storage management program 110a judges that it is unnecessary to create the storage domain 301 anew, and proceeds to S1010. On the other hand, if the allocation storage domain 301 is not present, the storage management program 110a judges that it is necessary to create the storage domain 301 anew, and proceeds to S1008 and S1009.

0098In S1008 and S1009, the storage management program 110a creates the storage domain 301 and adds and updates information on the created storage domain 301 in the storage domain management table 113. Thereafter, the storage management program 110a informs the storage subsystem 102 of updated or added information on the storage domain 301. The storage subsystem 102, which has receives the information, adds or updates the various tables included in the storage subsystem 102 on the basis of the received information. In this case, the storage subsystem 102 updates information on the various table to thereby associates respective ports included in one storage domain and respective logic devices using the updated information on the storage domain 301. Therefore, when the host computer 101 uses the logic devices included in the one storage domain, the host computer 101 only has to inform the management computer 103 of the storage domain (designate the storage domain). In the case in which a command or the like is received in a port included in one storage domain, if the command is for access to a logic device included in the same storage domain, the storage subsystem 102 performs processing such as reading or the like of data, and if the command is for access to a logic device included in a different storage domain, rejects the access. Note that the storage subsystem 102 may directly receive increase or the like of a port from the administrator or the like and update the various tables included in the storage subsystem 102. In this case, contents of the various tables included in the management computer 103 and contents of the various tables included in the storage subsystem 102 are different. In this case, as described later, the management computer 103 receives differential data from the storage subsystem 102 and updates the contents of the various tables included in the management computer 103. In addition, the storage management program 110a also adds and updates information on storage domains in the port management table 112. On the other hand, if the storage management program 110a has failed in the creation of the storage domain 301, the storage management program 110a reports an error and ends the processing (S1013).

0099Subsequently, the storage management program 110a allocates the device 105 to the allocation storage domain 301 (S1010). More specifically, the storage management program 110a stores an allocation storage domain ID in the entry 607 of the device management table 111. If the storage management program 110a has succeeded in the allocation of the device 105 to the storage domain 301, the storage management program 110a reports that the allocation is successful. If the storage management program 110a has failed in the allocation of the device 105 to the storage domain 301, the storage management program 110a reports an error (S1011 to S1013).

0100Next, port allocation processing, which is performed by the information processing system according to this embodiment, will be described using Figs. 11 to 15.

0101Fig. 11 is a diagram for explaining a flow of the port allocation processing according to this embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114), whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a and the name management program 114) are assumed to be subjects of operation.

0102In this embodiment, allocation of ports is performed by association of the name management program 114 and the storage management program 110a.

0103In addition, in this embodiment, port information of iSCSI with respect to the logic devices 105 is determined by the storage management program 110a when the storage management program 110a requests discovery. Therefore, the port information (Portal Group, Portal IP Address: Port) of the entries 805 and 806 in the name management table 115 is unregistered until discovery is requested for the logic devices 105.

0104Therefore, the name management program 114 registers the port information of the entries 805 and 806 in the name management table 115 after receiving a discovery request. Note that the name management program 114 stores information on the host computer 101, which includes port information, in the name management table 115 first.

0105In this embodiment, processing for carrying out the port allocation processing is roughly divided into two, namely, "configuration information update processing" and "discovery processing".

0106First, the configuration information update processing will be explained.

0107The storage management program 110a of the management computer 103 performs the "configuration information update processing". More specifically, the storage management program 110a performs processing for acquiring statistical information and configuration information from the storage subsystem 102 and the host computer 101 that are managed by the storage management program 110a. In the case in which update of configuration information is performed, the storage management program 110a carries out update processing for the device management table 111 and the port management table 112 (processing A 1101).

0108Note that the acquisition of configuration information performed by the storage management program 110a may be carried out periodically, or when information update is performed, the host computer 101 or the storage subsystem 102 may inform the storage management program 110a that the configuration information is updated. The storage management program 110a maintains latest information in this way.

0109Next, the "discovery processing" , which is performed by the name management program 114, will be described. The name management program 114 receives a request for discovery from the host computer 101. Here, in iSNS used in this embodiment, "DevAttrQry" corresponds to the discovery request. Upon receiving the discovery request, the name management program 114 carries out the discovery processing (processing B1102).

0110In the step of the port allocation processing in the discovery processing (processing B1102), the name management program 114 requests the storage management program 110a to perform allocation processing for the ports 204. The storage management program 110a, which has received the request, sends a list of available ports to the name management program 114. The name management program 114 updates the name management table 115 and, then, sends a result of allocation (list of ports) to the host computer 101.

0111In this way, in this embodiment, the name management program 114 can acquire the latest information from the storage management program 110a when the port allocation processing is performed. Therefore, according to this embodiment, port allocation, on which an operation state such as load and failure is reflected accurately, can be performed.

0112Next, the "configuration information update processing" will be described in detail using Fig. 12.

0113Fig. 12 is a diagram for explaining a flow of the configuration information update processing that is performed by the information processing system according to this embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114) or the CPU 202 of the storage subsystem 102 executes the storage management program 110b, whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a, the storage management program 110b, and the name management program 114) are assumed to be subjects of operation.

0114In starting the management computer 103, the storage management program 110a acquires the configuration information from the storage subsystem 102 and performs initialization of information in the respective table on the basis of the acquired configuration information (S1201).

0115Thereafter, the storage management program 110a acquires configuration information from the storage subsystem 102 under the management thereof at a predetermined interval (S1202). More specifically, the storage management program 110a inquires presence or absence of change in the configuration information (information on addition of the ports 204, information on a utilization rate of the ports, etc.) from the storage management program 110b of the storage subsystem 102. The storage management program 110b of the storage subsystem 102, which has received the inquiry, sends differential data to the storage management program 110a. Then, the storage management program 110a receives the differential data sent by the storage management program 110b of the storage subsystem 102.

0116The storage management program 110a judges whether or not the configuration of the storage subsystem 102 has been changed on the basis of the received differential data. If there is no data indicating the change in the configuration of the storage subsystem 102 (addition of the ports 204, etc.) in the differential data, the storage management program 110a judges that the configuration of the storage subsystem 102 has not been changed, and proceeds to S1205. On the other hand, if there is data indicating the change in the configuration of the storage subsystem 102 (addition of the ports 204, etc.) in the differential data, the storage management program 110a judges that the configuration of the storage subsystem 102 has been changed, and proceeds to step S1204 (S1203).

0117In step S1204, if the configuration has been changed, for example, increase or decrease from present configuration information, recovery from a blocked state, blocked from a normal state, or the like has occurred, the storage management program 110a registers new configuration information in corresponding entries in the device management table 111a, the port management table 112a, and the storage domain management table 113a.

0118In the case in which there is a port added anew, the storage management program 110a adds the port in the port management table 112, sets a storage domain ID for the port, and sets a state of the port as "normal". Note that the storage management program 110a registers data of addition, change, deletion, and the like of a storage domain ID in the corresponding respective tables in response to input from the administrator.

0119Next, the storage management program 110a checks an operation state of the ports 204. More specifically, the storage management program 110a judges whether or not there is the overload (high-load) port 204 out of data indicating utilization rates of the ports 204 in the differential data acquired in S1202. If there is data exceeding a "threshold value" decided in advance in the utilization rates acquired in S1202, the storage management program 110a proceeds to S1206. On the other hand, if there is no data exceeding the "threshold value" decided in advance in the utilization rate acquired in S1202, the storage management program 110a proceeds to S1207 (S1205).

0120In S1206, the storage management program 110a updates the entry 505 of the port 204 with the utilization rate exceeding the "threshold value" in the port management table 112 to "overload" (S1206). In this way, the port 204, for which "overload" is stored in the entry 505 of the port management table 112, is excluded from candidates of search in port allocation processing to be described later.

0121On the other hand, even if "overload" is stored in the port management table 112 for the port 204, if the acquired utilization rate for the port 204 is lower than the predetermined "threshold value", the storage management program 110a returns the corresponding entry 505 to "normal" (S1207 and S1208). Moreover, in S1207 and S1208, if there is the unregistered and low-load port 204, the storage management program 110a registers the port 204 in the port management table 112.

0122Note that, in this embodiment, the storage management program 110a repeats the processing of S1202 to S1208 after waiting for a predetermined time. Note that here, the storage management program 110a repeats the processing every fixed time. However, if a configuration or an operation state has been changed, delay in update of information may be prevented by applying trigger from the storage management program 110b to the storage management program 110a.

0123Next, the "discovery processing (processing B1102)" will be described in detail using Fig. 13.

0124Fig. 13 is a diagram for explaining a flow of the discovery processing that is performed by the management computer according to this embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114), whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a and the name management program 114) are assumed to be subjects of operation.

0125The name management program 114 receives a request for discovery (in "iSNS", "DevAttrQry") from an apparatus in the SAN 104, for example, the host computer 101 (S1301).

0126After receiving the request for discovery, the name management program 114 analyzes contents of the request and judges whether or not the request is related to the port allocation processing (S1302). If the request is not related to the port allocation processing, the name management program 114 proceeds to processing of S1305 and S1306. On the other hand, if the request is related to the port allocation processing, the name management program 114 proceeds to S1303.

0127In S1305 and S1306 , the name management program 114 creates a return message with reference to the device management table 111 and the port management table 112 in usual processing of the name management processing (i.e., iSNS) and, then, sends the return message to the host computer 101 that has requested the discovery.

0128Next, processing in S1303, to which the name management program 114 proceeds when the name management program 114 has received the request for the port allocation processing in S1302 , will be described.

0129In S1303, the name management program 114 requests the storage management program 110a to perform the port allocation processing (processing C1103 shown in Fig. 11). In requesting the port allocation processing (processing C1103), the name management program 114 adds information on the host computer 101 that has requested the processing ( "iSCSI Name" in the entry 802 of the name management table 115, "DD" in the entry 803, "Portal IP Address: Port" in the entry 806, etc. ) and information on an iSCSI node that is requested to be searched ( "iSCSI Name" in the entry 503 of the port management table 112) (S1303).

0130On the other hand, the storage management program 110a receives the request for the port allocation processing. Then, the storage management program 110a executes the port allocation processing and returns a result of the execution to the name management program 114 (the port allocation processing will be explained with reference to Fig. 14 later). The name management program 114 receives the result of the execution sent by the storage management program 110a (S1304).

0131Thereafter, the name management program 114 updates the name management table 115 on the basis of the received contents, creates a return message, and sends the return message to the request source host computer 101 (S1304 to S1306).

0132In this embodiment, in the processing of S1304 to S1306, the name management program 114 stores port information on the devices 105 in the name management table 115 (iSCSI Name 802, DD 803, Portal IP Address: Port 80.6, etc.).

0133Next, the "port allocation processing (processing C1103) will be described in detail using Fig. 14.

0134Fig. 14 is a diagram for explaining a flow of the port allocation processing that is performed by the management computer 102 according to this embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114), whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a and the name management program 114) are assumed to be subjects of operation.

0135The storage management program 110a receives a request for the port allocation processing sent by the name management program 114 (S1401).

0136The storage management program 110a refers to the device management table 111 to judge whether or not the requested logic device 105 (i.e., iSCSI node) is available (S1402) . If it is judged that the logic device 105 is available, the storage management program 110a proceeds to S1403.

0137On the other hand, if it is judged that the logic device 105 cannot be used, the storage management program 110a proceeds to S1409. Note that, the case in which the logic device 105 cannot be used includes a case in which the logic device 105 corresponding to the device management table 111 is not present and a case in which a "state" stored in the corresponding entry 506 is "blocked".

0138If it is judged in S1402 that the logic device 105 is available, the storage management program 110a searches the available storage domain 301 on the basis of the judged logic device 105 and information on the request source host computer 101, which is given when the name management program 114 requested the processing (S1403). More specifically, the storage management program 110a refers to the host management table 116 on the basis of a "device ID (i. e., iSCSI Name)" of the host computer 101 and a device ID of the logic device and acquires a storage domain ID to which the request source host computer 101 and the requested logic device belong.

0139If the appropriate storage domain 301 is present as a result of the search, the storage management program 110a can obtain one or more storage domain IDs. If the appropriate storage domain 301 is not present, the storage management program 110a sets a default storage domain ID. In the case in which the appropriate storage domain could be acquired and in the case in which the default storage domain ID could be set, the storage management program 110a proceeds to S1405. On the other hand, in the case in which the setting of the default storage domain ID is impossible, the storage management program 110a proceeds to S1409 (S1404).

0140In S1405, the storage management program 110a executes search processing for the obtained one or more storage domain IDs. The storage management program 110a searches through the port management table 112 to find the port 204 with a state of "normal" out of the ports 204 belonging to the respective storage domains 301 (S1405).

0141If the "normal" port 204 is not present in any of the storage domains 301, the storage management program 110a proceeds to S1409. If the "normal" port 204 is present, the storage management program 110a proceeds to s1407 (S1406).

0142In S1407, the storage management program 110a creates aport list. More specifically, in the case in which the "normal" port 204 is present in the plural storage domains 301 or in the case in which plural ports are allocated to the storage domain 301, the storage management program 110a calculates available throughput and the like from the number of "normal" ports 204 and a port utilization rate and selects a combination of the storage domains 301 and the ports 204 (group of the ports) that minimizes load. When the group of the ports 204 is selected, the storage management program 110a creates a port list from the selected group of the ports (S1407).

0143The storage management program 110a sends the created port list to the name management program 114 to end the processing (S1408).

0144If the storage management program 110a proceeds to S1409 in S1402, S1404 , and S1406, the storage management program 110a returns data indicating an error to the name management program 114 to end the processing.

0145Note that, in this embodiment, an order of the ports in the port list is not specifically limited. For example, the ports in the port list may be arranged in an order of search and output. However, since a port at the top of the port list is usually used first, it is advisable to arrange the port 204 with a highest priority at the top of the port list, and the port 204 with a lowest priority is arranged at the end of the port list. Note that information on performance and paths of the ports 204 may be used for the priorities of the ports in the port list.

0146For example, in the case in which priorities for port selection is selected according to performance, a port having a lowest utilization rate or having a margin in throughput (i.e., having large amount of unused throughput) in a result of the selection is given a highest priority.

0147In addition, in the case in which the closest port 204 is allocated on the basis of an address, for example, in the case of an IP network, sub-networks of IP addresses may be compared. A command for tracing a network path (in the Unix environment, there is a traceroute command) may be used to determine priorities taking into account paths from the host computer 101 (e.g., a priority for the port 204 with a short path from the host computer 101 is set high).

0148Moreover, as a method of determining priorities of the ports in the port list, a round robin method is possible. The storage management program 110a maintains priorities for the respective ports 204 and creates a port list with the port 204 having the highest priority at a top of the list. After sending the port list, the storage management program 110a changes the priority of the top port 204 to the lowest and sequentially moves up the priorities for the other ports 204. This makes it possible to access all the ports 204 uniformly with simple processing.

0149Next, a second embodiment of the present invention will be explained. In the second embodiment, in addition to the management for the storage domain 301 in the first embodiment, management for the storage subsystem is performed with the addition of topology management for the SAN 104. Note that, in the explanation of the second embodiment, components same as those in the first embodiment are denoted by the same reference numerals and signs.

0150Fig. 15 is a block diagram showing an example of a system configuration of an information processing system according to the second embodiment.

0151As shown in the figure, the information processing system according to the second embodiment has the same configuration as the information processing system according to the first embodiment except that a structure of the management computer 103 is different. In the following explanation, the different part will be mainly explained.

0152More specifically, in the management computer 103 according to the second embodiment, a topology management program 1501, the topology management table 1502, and the SAN domain management table 1503 are added to the management computer 103 according to the first embodiment. In addition, in the management computer 103 according to the second embodiment, a port management table 1504 and a host management table 1505 are provided instead of the management table 112 and the host management table 116 included in the management computer 103 according to the first embodiment.

0153In the port management table 1504 and the host management table 1505, information on a SAN domain is added to the management table 112 and the host management table 116 explained in the first embodiment.

0154The topology management program 1501 is a program that manages topology of the SAN 104 and security of Zoning, VLAN (Virtual LAN) , VPN (Virtual Private Network), and the like on a network. The information on topology is stored in the topology management table 1502.

0155In the storage management program 110a according to the second embodiment, a program for managing topology information is added to the storage management program 110a according to the first embodiment.

0156The CPU 10 of the management computer 103 executes the topology management program 150 to store topology information in a predetermined area in the memory 20 every time topology concerning the SAN 104 is changed. In the case in which the topology information is stored in the predetermined area in the memory 20, the CPU 10 executes the storage management program 110a to read out the topology information stored in the memory 20 and perform processing for updating the SAN domain management table 1503 and the port management table 1504 (or registering the information in the SAN domain management table 1503 and the port management table 1504). In this way, the CPU 10 executes the storage management program 110a and the topology management program 1501 in association with each other, whereby, every time the topology concerning the SAN 104 is changed, contents of the change can be reflected on the SAN domain management table 1503 and the port management table 1504.

0157In this way, in the second embodiment, the topology management for the SAN 104 is combined with the management for the storage domain 301. Such a configuration makes efficient path management possible in addition to the effects of the first embodiment. More specifically, this makes association with a security function or the like according to Zoning, VLAN (Virtual LAN), and VPN (Virtual Private Network) possible.

0158Fig. 16 is a diagram for explaining a SAN domain that is a management unit for the information processing system according to the second embodiment.

0159SAN domains 1601 shown in the figure are created by the CPU 10 of the management computer 103. More specifically, the CPU 10 executes the storage management program 110a to read out "information on topology and security of the SAN 104" stored in the predetermined area in the memory 20 as described above. The CPU 10 executes the storage management program 110a and uses the read-out "information on topology and security" to set the SAN domains 1601.

0160That is, the CPU 10 executes the storage management program 110a to set a group of ports, which are accessible in the SAN 104, as SAN domains on the basis of the security information. Next, the CPU 10 executes the storage management program 110a to classify the SAN domains 1601 in an identical security attribute in SAN domains 1601a to 1601n with respect to topology.

0161In the example shown in the figure, the CPU 10 executes the storage management program 110a to create the SAN domain 1601a first on the basis of a security attribute. Next, the CPU 10 executes the storage management program 110a to divide the created SAN domain 1601a into SAN domains 1601aa and 1601ab on the basis of the acquired topology information.

0162In this example, the host computer 101a belongs to both the SAN domains 1601aa and 1601ab, and the host computer 101b belongs to only the SAN domain 1601ab. In other words, although the logic device 105a and the logic device 105b are accessible from the host computer 101b, the access is possible only through the port 204b. On the other hand, the logic device 105a and the logic device 105b are accessible from the host computer 101a from both the ports 204a and 204b.

0163In addition, the example shown in the figure indicates that the SAN domains 1601aa and 1601ab belong to the storage domain 301a, and the SAN domain 1601c belongs to the storage domain 301b.

0164Next, data structures of the SAN domain management table 1503, the port management table 1504, and the host management table 1505 included in the information processing system according to the second embodiment will be explained using Figs. 17 to 19.

0165Fig. 17 is a diagram schematically showing an example of a data structure of the SAN domain management table 1503 according to the second embodiment.

0166The SAN domain management table 1503 includes an entry 1701 in which a "SAN domain ID" serving as an identifier for a SAN domain is stored, an entry 1702 in which a "SAN domain name" serving as a name of the SAN domain is stored, and an entry 1703 in which "attribute information" indicating attributes of the SAN domain is stored. The "SAN domain name" and "attribute data" are stored in the entries 1702 and 1703 in association with the "SAN domain ID" stored in the entry 1701.

0167In the entry of a top "row" in the SAN domain management table 1503 shown in the figure, data with the "SAN domain ID" of "na", the "SAN domain name" of "SAN domain A", and the "attribute data" of "VLAN-A" is stored.

0168The CPU 10 of the management computer 10 executes the storage management program 110a to therebymanage the data stored in the SAN domain management table 1503.

0169Fig. 18 is a diagram schematically showing a data structure of the port management table 1504 according to the second embodiment.

0170In the management port table 1504, an entry 1801, in which a "SAN domain ID" serving as an identifier for a SAN domain, is added to the port management table 112 shown in Fig. 5.

0171The CPU 10 of the management computer 10 executes the storage management program 110a to thereby manage data stored in the port management table 1504. The CPU 10 executes the storage management program 110a to thereby read out the topology information (information on topology and security of the SAN 104) stored in the predetermined area in the memory 20. The CPU 10 executes the storage management program 110a to analyze the read-out information and associate SAN domains with respective ports. In addition, the CPU 10 executes the storage management program 110a to store the associated SAN domains in the corresponding entry 1801.

0172Fig. 19 is a diagram schematically showing an example of a data structure of the host management table 1505 according to the second embodiment.

0173In the host management table 1505, an entry 1901, in which a "SAN domain ID" serving as an identifier for a SAN domain is stored, is added to the host management table 116 shown in Fig. 7.

0174The CPU 10 of the management computer 103 executes the storage management program 110a to thereby manage data stored in the host management table 1505. The CPU 10 executes the storage management program 110a to thereby read out the topology information (information on topology and security of the SAN 104) stored in the predetermined area in the memory 20. The CPU 10 executes the storage management program 110a to analyze the acquired information and associate SAN domains with respective ports. In addition, the CPU 10 executes the storage management program 110a to store the associated SAN domains in the corresponding entry 1901.

0175Fig. 20 is a diagram for explaining a flow of initial setting processing that is performed by the information processing system according to the second embodiment. Note that the initial setting processing, which is performed by the information processing system according to the second embodiment, is basically the same as the initial setting processing according to the first embodiment explained in Fig. 9. Thus, in the following explanation, different parts will be mainly explained.

0176The information processing system according to the second embodiment performs the same processing as S901 shown in Fig. 9 to receive setting for the storage domains 301 from an administrator (S2001).

0177Subsequently, in the second embodiment, the CPU 10 of the management computer 103 executes the storage management program 110a to set the SAN domain IDs 1801 in the port management table 1504 (S2002). In this embodiment, a specific procedure for the setting of the SAN domain IDs 1801 in the port management table 1504, which the CPU 10 performs by executing the storage management program 110a, is not specifically limited. For example, the CPU 10 may receive setting information for SAN domains inputted by the administrator and store the received setting information for the SAN domains in the port management table 1504.

0178Thereafter, the CPU 10 of the management computer 103 performs the same processing as S902 in Fig. 9 (S2003).

0179Next, device creation processing, which is performed by the information processing system according to the second embodiment, will be explained. The device creation processing, which is performed by the information processing system according to the second embodiment, is the same as the device creation processing according to in the first embodiment shown in Fig. 10.

0180Next, port allocation processing will be explained. The port allocation processing, which is performed by the information processing system according to the second embodiment, is the same as the port allocation processing according to the first embodiment except the port allocation processing (processing C) shown in Fig. 14. In other words, in the second embodiment, processing same as the processing according to the first embodiment shown in Figs. 1 to 13 is performed. Thus, the port allocation processing different from the port allocation processing according to the first embodiment will be explained using Fig. 21.

0181Fig. 21 is a diagram for explaining a flow of the port allocation processing that is performed by the management computer 103 according to this embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a, the name management program 114, and the topology management program 1501), or the CPU 202 of the storage subsystem 102 executes the storage management program 110a, whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a, the name management program 114, and the topology management program 1501) are assumed to be subjects of operation.

0182The storage management program 110a according to the second embodiment performs processing that is basically the same as the processing shown in Fig. 14. However, the processing is different from the processing shown in Fig. 14 in that processing for the SAN domains 1601 is added after S1404 shown in Fig. 14. In the following explanation, parts different from the processing in Fig. 14 will be mainly explained.

0183The storage management program 110a performs the same processing as S1401 to S1404. Thereafter, the storage management program 110a proceeds to processing of S2101 in which the storage management program 110a performs processing for the SAN domains 1601.

0184In S2101, the storage management program 110a also searches the SAN domain 1601 in the same procedure as the search for the storage domain 301. The storage management program 110a searches the host management table 1505 and obtains the SAN domain 1601 to which the request source host computer 101 belongs. In this case, if no SAN domain 106 is obtained, the storage management program 110a proceeds to S1409. On the other hand, if one or more SAN domain 1601 is obtained, the storage management program 110a proceeds to S2103 (S2102).

0185In S2103, the storage management program 110a searches through the port management table 1504 to acquire the port 204 with the state 605 of "normal" among the ports 204 belonging to both the storage domains 301 obtained in S1403 and the SAN domain 1601 obtained in S2101, and proceeds to processing of S1407.

0186On the other hand, if the port 204 with the state 605 of "normal" among the ports belonging to both the storage domains 301 obtained in S1403 and the SAN domain 1601 obtained in S2101 could not obtained as a result of the search, the storage management program 110a proceeds to S1409.

0187Note that, since processing in S1407 to S1409 is the same as the processing shown in Fig. 14, a description of the processing will be omitted.

0188In this way, in the second embodiment, the management for the storage domains 301 and the topology management for the SAN 104 are combined. Such a configuration makes efficient path management possible in addition to the effects of the first embodiment. More specifically, this makes association with a security function or the like according to Zoning, VLAN (Virtual LAN),and VPN(Virtual Private Network)possible. Here, association means, for example, dividing a SAN such that SAN domains are identical with domains in the VLAN.

0189Next, a third embodiment of the present invention will be explained. The third embodiment has the same configuration as the first and the second embodiments except that a procedure of the port allocation processing is different. Note that the third embodiment can be applied to both a configuration in which storage domains are set to manage the storage subsystem 102 as in the first embodiment and a configuration in which the storage subsystem 102 is managed using storage domains and SAN domains as in the second embodiment. In the following explanation, an example in which only storage domains are set to manage the storage subsystem 102 will be explained. In addition, in the explanation of the third embodiment, components same as those in the first embodiment are denoted by the same reference numerals and signs.

0190A configuration of the information processing system according to the third embodiment is the same as the configuration in the first embodiment shown in Figs. 1 and 2. However, the third embodiment is different from the first embodiment in the structures of the name management program 114 and the storage management program 110.

0191The storage domain 301, which is a management unit for the storage subsystem 102 according to the third embodiment, is the same as the storage domain 301 shown in Fig. 3.

0192Fig. 22 is a diagram for explaining port allocation processing that is performed by the management computer 103 according to the third embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114), whereby the port allocation processing explained below is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a and the name management program 114) are assumed to be subjects of operation.

0193In Fig. 22, a diagram for explaining a connection relation between the ports 204 and the logic devices 105 included in the storage subsystem 102 before changing the name management table 115 is shown in (a), and a diagram for explaining a connection relation between the ports 204 and the logic devices 105 included in the storage subsystem 102 after changing the name management table 115 is shown in (b).

0194In addition, (a-1) and (b-1) in Fig. 22 show the name management tables 115. Note that, in the name management tables 115 in (a-1) and (b-1), only data (the data of the entry 802 and the entry 805 shown in Fig. 8) necessary for explanation of the port allocation processing, which is performed by the management computer 103 according to the third embodiment, is extracted and illustrated (the name management table 115).

0195Further, in (a-2) and (b-2) in Fig. 22, images illustrating a connection relation between the ports 204 and the logic devices 105 included in the storage subsystem 102 are shown.

0196The name management table 115 shown in (a-1) shows a state in which a "logic device A" is connectable to the port 204a and the port 204b. In addition, the name management table 115 shown in (b-1) shows a state in which a "logic device B" is connectable to the port 204a and the port 204b.

0197In (a-2), the logic devices 105A and 105B and the ports 204a and 204b are allocated to the storage domain 301a. Broken lines shown in the figure indicate that the logic devices 105 and the ports 204 are connectable. In addition, (a-2) indicates that the device 105A is used through the port 204a, and the port 204a is in a state of "high-load". In the information processing system according to the third embodiment, as shown in (a-2), in the case in which the "port 204a" is in a high load state, new connection to the "port 204a" is limited.

0198The storage management program 110a according to the third embodiment monitors states of the ports 204 of the storage subsystem 102. If it is judged, as a result of monitoring, that the ports 204 are "overloaded", the storage management program 110a performs connection control such that new connection to the "port 204a" is not performed. More specifically, the storage management program 110a informs the storage subsystem 102 that use of the port 204a is to be stopped (or use of the port 204a is to be limited). When a port list is created (or after the port list is created), the name management program 114 prevents the port 204a frombeing included in the list.

0199However, in performing the connection control, the storage management program 110a secures at least one access path to the logic devices 105. In the case in which there is the logic device 105 having only one connection destination port 204, even if the connection destination port 204 is in the "overload" state, the storage management program 110a does not perform the connection control for the logic device 105. On the other hand, in the case in which the logic device 105 connectable to the "overload" port is also connectable to the ports 204 other than the overload port 204, the storage management program 110a controls connection to the logic device 105. More specifically, the storage management program 110a cancels a connectable relation between the logic device 105 and the "overload" port 204.

0200Note that, in the example shown in the figure, the device 105b has a connection path from the "port 204b" in addition to a connection path from the "port 204a". Thus, a connection path through the "port 204a" may be deleted. The storage management program 110a requests the name management program 114 to delete the "port 204a" form the connectable ports 204 of the device 105b. In this case, the storage management program 110a rejects an access request to the device 105b via the port 204a.

0201In response to the request, the name management program 114 changes the name management table 115 shown in (a-1) to the configuration shown in (b-1). In other words, the name management program 114 deletes data indicating the "port 204a" associated with the "device 105b" in the name management table 115. As a result, the connection relation between the ports 204 and the logic device 105 is as shown in (b-2).

0202In this way, in the third embodiment, in the storage domains 301, the ports 204 and the logic devices 105 are further associated with each other, and the connection relation between the ports 204 and the logic devices 105 is managed. The connection relation between the associated ports 204 and the logic devices 105 is changed according to loads of the ports 204.

0203That is, in the third embodiment, connection to the ports 204 is controlled according to a load state of the ports. This makes it possible to avoid access to the port 204 in a high-load state, and a utilization rate of the entire system can be improved. Note that data structures of the respective tables (the name management table 115, the host management table 116, the device management table 111, the port management table 112, and the storage domain management table 113) in the third embodiment are the same as those shown in Figs. 4 to 8.

0204In addition, initial setting processing and logic device creation processing, which are performed by the information processing system according to the third embodiment, are the same as those shown in Figs. 9 and 10.

0205Next, port allocation processing of the information processing system according to the third embodiment will be explained using Fig. 23. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114), or the CPU 20 of the storage subsystem 102 executes the storage management program 110b, whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a, the storage management program 110b, and the name management program 114) are assumed to be subjects of operation.

0206In performing configuration information update processing, the storage management program 110a of the management computer 103 inquires configuration information from the storage management program 110b of the storage subsystem 102 under the control of the storage management program 110a. The storage management program 110a acquires configuration information that is sent by the storage management program 110b in response to the inquiry.

0207In the case in which a notice of change for a configuration to the name management program 114 is necessary in the case of a change in a configuration, detection of a high-load port, shift from high-load to low-load, or the like, the storage management program 110a issues a request for registration change for configuration information to the name management program 114 (processing D2301).

0208The name management program 114 applies the same processing as the discovery request shown in Fig. 11 to a discovery request from the host 101.

0209Fig. 24 is a diagram for explaining a flow of configuration update processing that is performed by the information processing system according to this embodiment. Note that the CPU 10 of the management computer 103 executes the respective programs (the storage management program 110a and the name management program 114), or the CPU 202 of the storage subsystem 102 executes the storage management program 110b, whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110a, the storage management program 110b, and the name management program 114) are assumed to be subjects of operation.

0210In the third embodiment, in the case in which a configuration of the storage subsystem 102 is updated, the storage management program 110a is required to inform the name management program 114 of contents of the update and reflect the contents of the update on the name management table 115. In the third embodiment, a change list table (not shown) is provided in the memory 20 of the management computer 103. The storage management program 110a stores contents of a change in the change list table. More specifically, the information processing system according to the third embodiment performs processing as described below.

0211The storage management program 110a performs the same processing as S1201 to S1204 shown in Fig. 12 and proceeds to processing in S2401. In S2401, the storage management program 110a stores data concerning the configuration change in the change list table, which is prepared in the memory 20 in advance, and proceeds to the processing in S1205. Thereafter, the storage management program 110a performs the same processing as S1205 and S1206 shown in Fig. 12 and proceeds to processing in S2402.

0212In S2402, the storage management program 110a stores data concerning an overload port in the change list and proceeds to the processing in S1207. Thereafter, the storage management program 110a performs the same processing as S1207 and S1208 shown in Fig. 12 and proceeds to processing in S2403.

0213In S2403, the storage management program 110a stores data concerning an unregistered and low-load port (or data concerning a port whose state has changed from high-load to low-load) in the change list and proceeds to processing in S2404.

0214In S2404, the storage management program 110a sends contents of the change list to the name management program 114 and performs the same processing as S1209 shown in Fig. 12.

0215In this way, according to the third embodiment, the storage management program 110a of the management computer 103 monitors states of the ports 204 of the storage subsystem 102 and performs connection control such that new connection is not made to the "overload" port 204. Therefore, according to the third embodiment, in addition to the effects of the first embodiment, access to the port 204 in an overload state can be avoided, and a utilization rate of the entire system can be improved.

0216Next, a fourth embodiment of the present invention will be explained. The fourth embodiment has the same configuration as the first and the second embodiments except that a procedure for port allocation processing is different. The fourth embodiment of the present invention can be applied to both the configuration in which storage domains are set to manage the storage subsystem 102 as in the first embodiment and the configuration in which storage domains and SAN domains are used to manage the storage subsystem 102 as in the second embodiment. In the following explanation, an example in which only storage domains are set to manage the storage subsystem 102 will be explained. In addition, in the explanation of the fourth embodiment, components same as those in the first embodiment are denoted by the same reference numerals and signs.

0217Fig. 25 is a block diagram showing an example of a system configuration of an information processing system according to the fourth embodiment.

0218The information processing system according to the fourth embodiment shown in the figure is the same as the information processing system according to the first embodiment except that a structure of the management computer 103 is different.

0219More specifically, at least the name management program 114 and the name management table 115 are stored in the memory 20 of the management computer 103. The CPU 202 of the storage subsystem 102 executes the storage management program 110b to send configuration information of the storage subsystem 102 to the management computer 103.

0220The CPU 10 of the management computer 103 executes the name management program 114 to receive the configuration information sent from the CPU 202 of the storage subsystem 102. The CPU 10 of the management computer 103 executes the name management program 114 to perform registration and update of data in the name management table 115 using the received configuration information.

0221Data structures of the respective tables (the name management table 115, the host management table 116, the device management table 111, the port management table 112, and the storage domain management table 113) in the fourth embodiment are the same as those shown in Figs. 4 to 8.

0222In addition, initial setting processing and logic device creation processing, which are performed by the information processing system according to the fourth embodiment, are the same as the processing shown in Figs. 9 and 10.

0223Fig. 26 is a diagram for explaining a flow of port allocation processing that is performed by the information processing system of the fourth embodiment. Note that the CPU 10 of the management computer 103 executes the name management program 114, or the CPU 202 of the storage subsystem 102 executes the storage management program 110b, whereby the following processing is realized. However, in the following explanation, for simplification of the explanation, the respective programs (the storage management program 110b and the name management program 114) are assumed to be subjects of operation.

0224The fourth embodiment adopts a form in which the storage management program 110b of the storage subsystem 102 directly communicates with the name management program 114. It is unnecessary to perform communication among the storage management programs 110 (between the storage management programs 110a and 110b) as in the first embodiment.

0225In the case in which there is a change in configuration information such as a change in a configuration, detection of a high-load port, or shift from high-load to low-load, the storage management program 110b issues a request for registration change for configuration information to the name management program 114 (S2301).

0226The processing in S2301 is the same as the processing shown in Fig. 12 except that the processing procedure in S1203 is different. More specifically, the processing in S2301 is different from the processing shown in Fig. 12 in that, in the case of S2301, rather than performing the update information inquiry from the storage management program 110a to the storage management program 110b as in S1203 in Fig. 12, the storage management program 110b refers to the configuration information of the storage subsystem 102.

0227Note that, in the case of a discovery request from the host computer 101, the name management program 114 performs ordinary processing for the discovery request and responds to the discovery request.

0228In this way, in the first to the fourth embodiments, the group of the logical devices 105, the group of the ports 204, and the host computers 101 are classified in the storage domain 301A such that logic devices and ports belonging to the same storage domain can be combined arbitrarily. Therefore, it becomes unnecessary to perform setting for associating logic devices and ports in a one-to-one correspondence as in the conventional storage subsystem, and complexity of management of the storage subsystem can be eliminated.

0229In addition, according to this embodiment, in performing the port allocation processing, ports can be allocated to logic devices dynamically taking into account load states of the ports. Further, according to this embodiment, in performing the port allocation processing, the ports 204, which have addresses near the host 101 requesting access, can be allocated dynamically. Therefore, according to this embodiment, a utilization rate of the entire system can be improved.

0230Moreover, in this embodiment, a host computer sends a standard discovery request complying with protocols of iSNS and iSCSI to a management computer to thereby acquire information on a port corresponding to a logic device that is desired to be accessed from the management computer. Consequently, according to this embodiment, in the case in which load distribution and failure avoidance processing for a information processing system, to which plural storage subsystems of different manufacturers are connected, are performed, it is unnecessary to install a device program for each of disk array apparatuses of the different manufacturers on a host computer.

0231Note that the present invention is not limited to the embodiments explained above, and various modifications are possible within a scope of the present invention. For example, in the third and the fourth embodiment, the case in which a storage domain is used as a management unit for a storage subsystem is explained. However, a SAN domain may be used as the management unit for the storage subsystem in addition to the storage domain as in the second embodiment.

0232In addition, in this embodiment, host computers, logic devices, and ports are grouped in storage domains to perform access management. However, the present invention is not limited to this. For example, it is also possible that only the logic devices and the ports are grouped in the storage domains, and the access management with respect to the logic devices is performed from the host computers. In the case of such a configuration, since it is also unnecessary to associate the ports and the logic devices one to one as in the conventional technique, time and labor at the time of setting for a storage subsystem and a change in the configuration can be reduced. If the logic devices and the ports are grouped in this way, since optimal ports can be allocated according to states of the ports, a utilization rate of the entire system can be improved as in the above-mentioned embodiments.

0233In this way, according to the present invention, plural ports, plural logic devices included in a storage subsystem, and host computers are grouped in one or more storage domains. Then, in the present invention, a storage domain to which both a host computer, which has sent a discovery request, and a logic device, for which the discovery is requested, belong is selected out of the storage domains in which the plural ports, the plural logic devices, and the host computers are grouped, and ports belonging to the specified storage domain are selected. Consequently, in the present invention, simply by grouping the plural ports and the plural logic devices, the ports can be allocated without setting the logic devices and the ports in one-to-one correspondence. In other words, according to the present invention, in the storage subsystem, interface ports can be managed and used easily and efficiently.

0234Further, according to the present invention, the host computer can acquire information on ports corresponding to a logic device, which is desired to be accessed, from a management computer by sending a standard discovery request complying with the protocols of iSNS and iSCSI to the host computer. Consequently, according to the present invention, even in the case in which load distribution and failure avoidance processing for an information processing system, to which plural storage subsystems of different manufacturers are connected, are performed, it is unnecessary to install a device program for the storage subsystem of the different manufacturers on a host computer.