WO2015022908A1 - Storage supply system, storage selection device, storage supply method, and program - Google Patents

Abstract

The present invention enables a host device that is connected to a network to access the most preferable storage. The storage supply system includes: two or more storages having different characteristics; a host device that accesses these storages; a switch located between the storages and the host device; a storage selection device that selects the storage to be used by the host device, on the basis of the characteristics of a service operating in the host device and the characteristics of the storages; and a control device that controls the switch so as to enable the host device to access the selected storage.

Description

Storage providing system, storage selection device, storage providing method, and program

(Description of related applications)
The present invention is based on the priority claim of Japanese patent application: Japanese Patent Application No. 2013-167791 (filed on August 12, 2013), the entire contents of which are incorporated herein by reference. Shall.
The present invention relates to a storage providing system, a storage selecting device, a storage providing method, and a program, and more particularly to a storage providing system, a storage selecting device, a storage providing method, and a program for selecting a storage and accessing a host device.

Patent Document 1 discloses a data migration technique that can satisfy required performance as much as possible. According to this document, the server 11 obtains IOPS (I / O per second) measurement values of each business AP (application) and IOPS measurement values of each logical volume (VOL) that is the access destination of each business AP. Based on the obtained various IOPS measurement values, a source sub candidate is selected from a plurality of VOLs. The server 11 discriminates a parity group to which the selected source sub-candidate belongs from among a plurality of parity groups, sets two or more VOLs belonging to the determined parity group as two or more source candidates, and sets a plurality of other VOLs Select each target candidate from the list. The server 11 selects a source candidate and target candidate pair for which the most preferable migration is performed and / or the result of the migration is estimated to be the most preferable, and the data is migrated from the source to the target constituting the selected pair. , And is described.

Patent Document 2 discloses a network device that realizes live migration processing between different bases without transferring data stored in a storage area to another base. According to this document, this network device includes a processor, a memory, and an interface, the memory stores a program for realizing the access processing unit and connection management information, and the access processing unit notifies the start of the live migration process. , Refer to the connection management information, identify the port for accessing the storage area allocated to the virtual machine, and acquire the address of the transfer destination device that transfers the access request sent from the virtual machine Then, it is described that the conversion information is generated by associating the port and the address, and the access request transmitted after the live migration process is executed is controlled based on the conversion information.

Storage connected to the network includes storage that stores data in various forms such as files, tables, key values, and raw. Further, even with the same type of storage, there are differences in access speed, capacity efficiency, cost, possible search methods, etc., depending on the type of storage. These storage types and types are determined at the time of application initialization. In addition, in ILM (Information Lifecycle Management), it is possible to move data between storages of the same type by using a storage virtualization device for access to the storage in consideration of the data access frequency. (See paragraphs 0007-0008 of Patent Document 1).

Non-Patent Documents 1 and 2 are examples of networks that can control an access path between a host and a storage.

JP 2007-156815 A JP 2013-30085

Nick McKeown and 7 others, “OpenFlow: Enabling Innovation in Campus Networks”, [online], [Search July 2, 2013], Internet <URL: http://www.openflow.org/documents/ openflow-wp-latest.pdf> “OpenFlow Specification” Version 1.3.1 (Wire Protocol 0x04), [online], [Search July 2, 2013], Internet <URL: https://www.opennetworking.org/images /stories/downloads/sdn-resources/onf-specifications/openflow/openflow-spec-v1.3.1.pdf>

The following analysis is given by the present invention. As described above, various types of storage may be connected to the network, but there is a problem that there is no guarantee that the host device connected to the network can access the optimal storage. For example, storage is selected at the time of initialization of an application of the host device, but the characteristics of the storage are not considered at all.

In Patent Document 1, at the time of data migration, the performance of the access source of each business AP (application) (access source performance) and the performance of each logical volume itself that is the access destination of each business AP (volume performance) are acquired. It describes that a pair of a source volume and a target volume is selected from a plurality of logical volumes based on each acquired performance value. However, this document does not describe the selection of a storage in consideration of the service running on the host, particularly its characteristics.

It is an object of the present invention to provide a storage providing system, a storage selection device, a storage providing method, and a program that can contribute to improving the accessibility to storage from a host device connected to a network.

According to the first aspect, two or more storages each having different characteristics, a host device that accesses the storage, a switch disposed between the storage and the host device, and an operation on the host device A storage selection device that selects a storage to be used by the host device based on the characteristics of the service to be performed and the characteristics of the storage; and a control device that controls the switch so that the host device can access the selected storage; , A storage providing system is provided.

According to a second aspect, two or more storages each having different characteristics, a host device that accesses the storage, a switch disposed between the storage and the host device, and a control that controls the switch A storage selection unit for selecting a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage, A storage selection device is provided that includes a notification unit that notifies the control device of a combination of the selected storage and a host device, and controls the switch so that the host device can access the selected storage.

According to a third aspect, two or more storages each having different characteristics, a host device that accesses the storage, a switch arranged between the storage and the host device, and a control that controls the switch A storage providing system to which the host device is connected, selecting a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage; and the selected storage; A storage providing method comprising: notifying the control device of a combination with a host device, and controlling the switch so that the host device can access the selected storage. This method is associated with a specific machine called a storage selection device that selects storage on a network.

According to a fourth aspect, two or more storages each having different characteristics, a host device that accesses the storage, a switch disposed between the storage and the host device, and a control that controls the switch A process for selecting a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage in a computer connected to the storage providing system to which the device is connected; There is provided a program for notifying the control device of a combination of the selected storage and the host device, and executing processing for controlling the switch so that the host device can access the selected storage. This program can be recorded on a computer-readable (non-transient) storage medium. That is, the present invention can be embodied as a computer program product.

According to the present invention, it is possible to contribute to improving the accessibility to the storage from the host device connected to the network.

It is a figure which shows the structure of the storage provision system of the 1st Embodiment of this invention. It is a figure which shows the detailed structure of the switch of the 1st Embodiment of this invention. It is a figure which shows the example of the control information (flow entry) set to the flow table of the switch of the 1st Embodiment of this invention. FIG. 5 is a diagram illustrating an example of actions (processing instructions) that can be set in an instruction field of control information (flow entry) in FIG. 4. It is a figure which shows the detailed structure of the control apparatus of the 1st Embodiment of this invention. It is a figure which shows the example of the entry hold | maintained at the host-storage corresponding | compatible information storage part which the control apparatus of the 1st Embodiment of this invention hold | maintains. It is a figure which shows the example of the entry of the topology database which the control apparatus of the 1st Embodiment of this invention hold | maintains. It is a figure which shows the detailed structure of the storage selection apparatus of the 1st Embodiment of this invention. It is a figure which shows the example of the entry of the storage characteristic database which the storage selection apparatus of the 1st Embodiment of this invention hold | maintains. It is a flowchart showing operation | movement of the switch of the 1st Embodiment of this invention. It is a flowchart showing operation | movement of the control apparatus of the 1st Embodiment of this invention. It is a flowchart showing operation | movement of the storage selection apparatus of the 1st Embodiment of this invention. It is a sequence diagram showing operation | movement of the storage provision system of the 1st Embodiment of this invention. It is another sequence diagram showing operation | movement of the storage provision system of the 1st Embodiment of this invention. It is a figure which shows the structure of the storage provision system of the 2nd Embodiment of this invention. It is a figure which shows the detailed structure of the storage selection apparatus of the 2nd Embodiment of this invention. It is a figure which shows the example of the entry hold | maintained at the use storage information storage part hold | maintained by the storage selection apparatus of the 2nd Embodiment of this invention. It is a figure which shows the detailed structure of the storage management apparatus of the 2nd Embodiment of this invention. It is a flowchart showing operation | movement of the storage selection apparatus of the 2nd Embodiment of this invention. It is a flowchart showing operation | movement of the storage management apparatus of the 2nd Embodiment of this invention. It is a sequence diagram showing operation | movement of the storage provision system of the 2nd Embodiment of this invention. It is a sequence diagram showing operation | movement of the storage provision system of the 2nd Embodiment of this invention. It is a figure which shows the structure of the storage provision system of the 3rd Embodiment of this invention. It is a figure which shows the structure of the storage provision system of the 4th Embodiment of this invention. It is a figure which shows the detailed structure of the storage selection apparatus of the 5th Embodiment of this invention.

First, an outline of an embodiment of the present invention will be described with reference to the drawings. Note that the reference numerals of the drawings attached to this summary are attached to the respective elements for convenience as an example for facilitating understanding, and are not intended to limit the present invention to the illustrated embodiment.

In one embodiment of the present invention, two or more storages (2a to 2n in FIG. 1) having different characteristics, a host device (1 in FIG. 1) that accesses the storage, the storage, and the host device And a storage selection device for selecting a storage to be used by the host device based on the characteristics of the service operating on the host device and the characteristics of the storage (3 in FIG. 1) 1) and a control device (4 in FIG. 1) that controls the switch so that the host device can access the selected storage.

The storage selection device (5 in FIG. 1) selects the storage 2a as the storage to be used by the host device 1 in FIG. 1 based on, for example, the characteristics of the service operating on the host device and the characteristics of the storage. And notify the control device (4 in FIG. 1). Based on the notification, the control device (4 in FIG. 1) controls the switch (3 in FIG. 1) to transmit the packet addressed to the storage from the host device 1 to the storage 2a. By doing so, the host device connected to the network can access more desirable storage.

[First Embodiment]
Next, a first embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a configuration of a storage providing system according to the first embodiment of this invention. Referring to FIG. 1, a switch 3, a host device 1 and storages 2a to 2n connected via the switch 3, a control device 4 for controlling the switch 3, and a storage to which the host device 1 is connected are selected. Thus, a configuration including a storage selection device 5 that notifies the control device 4 is shown.

The host device 1 is a computer in which application programs that use the storages 2a to 2n are installed. In the example of FIG. 1, one host device 1 is shown, but a plurality of host devices 1 may be provided. That is, in the present invention, it is possible to select an optimum storage for each of the m host apparatuses 1.

The storages 2a to 2n are storage devices such as a magnetic disk, a memory, an optical disk, a magneto-optical disk, an array device thereof, or a server device on which these are mounted.

The switch 3 is a switch that operates according to control information (flow entry) set from the control device 4. As such a switch, the open flow switch of the nonpatent literatures 1 and 2 can also be utilized. In the example of FIG. 1, one switch 3 is shown, but a plurality of switches 3 may be provided.

The control device 4 is a device that controls the switch 3. As such a control apparatus 4, the open flow controller of a nonpatent literature 1, 2 can also be utilized.

The storage selection device 5 includes service information describing how the application program operating on the host device 1 uses the storage (what type of storage is requested), and the characteristics of the storages 2a to 2n. Based on the above, a storage to be accessed by the host device 1 is determined and notified to the control device 4.

Subsequently, the configuration of the switch 3, the control device 4, and the storage selection device 5 will be described in detail.

FIG. 2 is a diagram showing a detailed configuration of the switch according to the first embodiment of the present invention. Referring to FIG. 2, a switch 3 having a packet processing unit 30 and a flow table 31 is shown.

The packet processing unit 30 is a logical means for processing a packet that has arrived at the switch 3, and is realized by a processor, a memory, or the like.

The flow table 31 corresponds to a rule (match condition) that is referred to by the packet processing unit 30 and collates with the header information of the received packet, and a process (action) that is performed on a packet that matches the rule (match condition). It is a table for storing attached control information (flow entry). As the flow table 31, CAM (Content Addressable Memory) can also be used.

FIG. 3 is a diagram illustrating an example of control information (flow entry) set in the flow table 31 of the switch according to the first embodiment of this invention. Referring to FIG. 3, a field for storing a rule (match condition) for collating with a packet header or the like of a received packet, and a flow statistics information field (Counters) for storing statistical information such as a packet matching the rule (match condition), , An entry in which an instruction field (Instructions) for storing processing contents (Actions) to be applied to a packet conforming to the rule (match condition) is associated is shown. Note that it is also possible to set a wild card as a rule (match condition).

FIG. 4 is a diagram showing an example of processing contents (Action) that can be set in the instruction field (Instructions) and the contents thereof. OUTPUT is an action for outputting a received packet to a designated port (interface). SET_VLAN_VID to SET_TP_DST are actions for modifying the field of the packet header. By combining these, for example, a packet addressed from a certain transmission source to a destination can be output from a designated port after rewriting the VLAN ID, for example. 3 and 4 are equivalent to the OpenFlow switch of Non-Patent Document 2, but other equivalent mechanisms can be used.

FIG. 5 is a diagram showing a detailed configuration of the control device according to the first embodiment of the present invention. Referring to FIG. 5, a control device 4 including a host-storage correspondence information storage unit 44, a route calculation unit 41, a topology database (topology DB) 42, and a flow entry setting unit 43 is shown.

The host-storage correspondence information storage unit 44 stores host-storage correspondence information indicating the correspondence relationship between the host device 1 and the storage to be used by the host device 1.

The route calculation unit 41 is a logical processing entity that calculates the route of the packet sent from the switch 3 with reference to the host-storage correspondence information storage unit 44 and the topology DB 42.

The topology DB 42 is a database that manages the connection relationship (topology) among the host device 1, the storages 2a to 2n, and the switch 3.

The flow entry setting unit 43 is a logical means for performing communication for setting an entry in the flow table 31 of the switch 3.

FIG. 6 is a diagram illustrating an example of entries held in the host-storage correspondence information storage unit 44. In the example of FIG. 6, an entry that associates host information, virtual storage information, and real storage information is shown. As the host information, for example, an IP (Internet Protocol) address of the host device 1, a MAC (Media Access Control) address, or both can be used.

The virtual storage information is storage information designated by the host device 1 as an access destination. For example, as shown in FIG. 6, an IP address (may be a MAC address), an IP protocol number (upper layer protocol), and a destination port address of TCP / UDP (Transmission Control Protocol / User Datagram Protocol) can be used. .

Real storage information is information on storage that is actually accessed by the host device 1. As the real storage information, an ID of the connected switch (switch ID; data path ID), a port of the switch, for example, a storage IP address, a storage port, and the like as shown in FIG. 6 can be used. Of course, the storage MAC address or VLAN (Virtual Local Area Network) tag may also be included.

FIG. 7 is a diagram illustrating an example of entries held in the topology DB 42. The upper part (a) of FIG. 7 is an entry indicating the connection relationship between the switches 3 and is represented by a set of ports connected to the switch ID (DPID).

The lower part (b) of FIG. 7 is an entry indicating a connection relationship between the switch 3 and the host device 1 or between the switch 3 and the storages 2a to 2n. In the example of FIG. 7, the switch 3 side is expressed by a switch ID (DPID) and its port, and the host device 1 or the storage 2 side is expressed by a MAC (Network Interface Card) MAC address.

FIG. 8 is a diagram showing a detailed configuration of the storage selection device according to the first embodiment of the present invention. Referring to FIG. 8, a storage selection device 5 including a service analysis unit 51 and a storage characteristic DB 52 is shown.

The service analysis unit 51 is a logical means for selecting a storage suitable for the host device 1 based on the service information and notifying the control device 4 of the result. The service analysis unit 51 includes a storage selection unit that selects a storage to be used by the host device 1 based on characteristics of a service operating on the host device 1 and characteristics of the storage, and the selected storages 2a to 2n and the host It also functions as a notification unit that notifies the control device 4 of the combination with the device 1 and controls the switch 3 so that the host device 1 can access the selected storage.

The storage characteristic DB 52 is a database storing information such as the type of each storage.

FIG. 9 is a diagram showing an example of entries held in the storage characteristic DB 52. The example of FIG. 9 shows an entry in which a storage ID for uniquely identifying a storage, a storage type, an in-type type, an attribute, storage information, and the like are associated with each other.

In the storage type field, the type of the corresponding storage, for example, file system, database, key-value store, RAW, etc. is set.

In the type type field, a sub-item of the type set in the storage type field, for example, a database engine name, a key-value store software name, a file system name, and the like are set.

In the attribute field, the type of storage, for example, capacity efficiency, capacity unit price, read speed, write speed, search speed, failure strength (redundancy), size suitability (suitable for large size, suitable for small size) Etc.) is set. In addition to setting absolute values for these items, for example, relative values such as high capacity efficiency, low capacity, and normal capacity may be set.

In the storage information field, as shown in the lower part of FIG. 9, to which switch the corresponding storage is connected (switch ID), to which port of the switch (switch port), and usage (capacity division and , An access method {IP address, port number, etc.} to the divided partition is set.

Here, the service information input to the service analysis unit 51 of the storage selection device 5 will be described. In the service information, storage requirements and the like required for the host device 1 to receive a service from the storage are set based on an application program or the like installed in the host device. Examples of items that are desirably set as service information are shown below.
Storage access information: IP address, port number, protocol, etc. Storage type: File system, database, key-value store, RAW, etc. Storage requirements: Required capacity, price, request read / write / search speed, request failure strength, data size The service information format includes, for example, XML (Extensible Markup Language), CSV (comma-separated values), and the like.

Of course, it is possible to provide a service information management device that displays a service information management screen in response to a request from the user terminal or the like, accepts input or editing of the service information, and provides the storage selection device 5 with it. Further, the storage selection device 5 may have a function corresponding to the service information management device.

Note that each unit (processing means) of the control device 4 shown in FIG. 1 can be realized by a computer program that causes a computer constituting the control device 4 to execute the above-described processes using the hardware.

Subsequently, the operation of the present embodiment will be described in detail with reference to the drawings. First, the basic operation of each device will be described in detail with reference to the drawings.

FIG. 10 is a flowchart showing the operation of the switch according to the first embodiment of the present invention. When the switch 3 receives a new packet from the host device 1 on which an application that uses the storage is activated, the switch 3 receives control information (flow entry) having a rule (match condition) that matches the header of the received packet from the flow table 31. ) Is searched (step S100).

When the corresponding control information (flow entry) is found as a result of the search (Yes in step S101), the switch 3 performs an instruction field (Instructions; see FIG. 3) of the control information (flow entry) for the received packet. The action set to is applied (step S102).

On the other hand, when the corresponding control information (flow entry) is not found in Step S100 (No in Step S101), the switch 3 transfers the received packet to the control device 4 (Step S103).

Next, the operation of the control device 4 that is started when the packet is transferred from the switch 3 will be described. FIG. 11 is a flowchart showing the operation of the control device according to the first embodiment of the present invention.

Referring to FIG. 11, when the control device 4 receives the packet transfer from the switch 3, the control device 4 searches the host-storage correspondence information storage unit 44 for an entry that matches the received packet (step S200). Here, the entry whose source IP address matches the IP address of the host information and whose destination IP address matches the IP address of the virtual storage information (see FIG. 6), or the source IP address etc. An entry that matches the IP address or the like of the actual storage information (see FIG. 6) and whose destination IP address matches the IP address of the host information is searched.

As a result of the search, when a corresponding entry is found (Yes in step S201), the control device 4 refers to the found entry and obtains information on the storage specified in the packet. At this time, if the storage specified by the packet is virtual storage, the control device 4 converts it to real storage information, and if the storage specified by the packet specifies real storage, converts it to virtual storage information ( Step S202). As a result, the host device 1 recognizes that it is accessing the storage of the virtual storage information, but the storage that is actually accessed is the storage specified by the real storage information. By doing so, it becomes possible for the host device 1 to access the optimum storage (selected by the storage selection device 5) without being aware of the storage that is actually accessed.

Next, the control device 4 uses the information in the topology DB 42 based on the packet header and, if storage information is obtained in step S202, the input / output switch information, the packet transfer path (multiple switches are If there is, the packet transfer path on these switches is calculated (step S203).

Next, the control device 4 calculates control information (flow entry) to be set in the flow table 31 of each switch based on the route calculated in step S203 (step S204). For example, when transferring from the host device 1 to the storage 2a, control information (flow entry) is calculated that causes the switch 3 to transfer a packet addressed to the virtual storage from the host 1 to the storage 2a specified by the real storage information. The

Next, the control device 4 sets the control information (flow entry) calculated in step S204 in the flow table 31 of the corresponding switch 3 (step S205).

Finally, in accordance with the entry obtained in step 201, the control device 4 sends the packet with the changed packet header to the destination host device or storage 2a to 2n to the destination host or storage connected switch. (Step S206; packet output instruction).

In step S204, the rule (match condition) of the control information (flow entry) to be calculated can be obtained as follows. First, when the packet transmission source is the connected switch 3 (in the case of a switch located at the start point of the packet transfer path), the rule (match condition) of the corresponding switch includes a content that matches the packet header before conversion. Is set. On the other hand, if it is a switch other than that (in the case of a switch located downstream from the starting point of the packet transfer path), the content that matches the packet header after being converted between the real storage information and the virtual storage information in step S202. Used as a rule (match condition).

The conversion process between the real storage information and the temporary storage information in step S202 described above is control information (flow entry) that rewrites the header of the packet to the switch to which the transmission source is connected (switch located at the start point of the packet transfer path). ) Can also be realized.

Subsequently, the operation of the storage selection device 5 for adding or changing entries in the host-storage correspondence information storage unit 44 will be described. FIG. 12 is a flowchart showing the operation of the storage selection device 5.

First, the storage selection device 5 starts analysis upon reception of input of service information and the like, and extracts the storage used by the application of each host device 1 and its requirements (step S301).

Next, the storage selection device 5 refers to the storage characteristic DB 52 (see FIG. 9) and searches for a storage that meets the requirements extracted in step S301 (step S302). At this time, when a plurality of storages are found, the storage selection device 5 selects an optimum storage from among them. As a criterion for selecting a storage, there is a method of comparing the attributes of each storage with the requirements extracted in step S301 and selecting the one with the most matching items. Of course, each item may be appropriately weighted to calculate a score or the like, or a storage selection criterion (selection policy) created in advance may be used. Alternatively, a storage load state or the like may be obtained separately, and for example, a storage with a small load at that time or in the future may be selected.

Next, the storage selection device 5 performs the operation of updating the storage information of the storage determined in step S301 in the storage characteristic DB 52 and initializing the corresponding storage as necessary (step S303).

Next, the storage selection device 5 gives the control device 4 information on the host device 1, virtual storage information that is storage information viewed from the host device 1, and real storage information that is actual storage information obtained from the storage characteristics DB 52. Information is notified (step S304). As a result, a new entry is added to the host-storage correspondence information storage unit 44 of the control device 4.

FIG. 13 is a sequence diagram showing the operation of the storage providing system according to the first embodiment of the present invention achieved by the operation of each device.

When the user inputs service information to the storage selection device 5 (step S1001), the storage selection device 5 selects the storage 2 (hereinafter referred to as “storage 2” if the storages 2a to 2n are not particularly distinguished). The storage 2 is initialized as necessary (steps S1002 and S1003).

Next, the storage selection device 5 notifies the control device 4 of the information of the host device 1, the virtual storage information of the selected storage, and the real storage information (step S1004). The control device 4 adds a new entry to the host-storage correspondence information storage unit 44 based on the notified contents (steps S1004 to S1006).

Thereafter, when the user activates the application (step S1007) and the host device 1 sends out a packet (read) requesting access to the storage 2 (step S1008), the switch 3 sends the packet (read) to the control device. 4 (step S1009).

The control device 4 extracts an entry corresponding to the transferred packet (read) from the host-storage correspondence information storage unit 44, and sets control information (flow entry) in the switch 3 (step S1010). Further, the control device 4 transmits the packet (read) received in step S1009 to the switch 3 (step S1011) and instructs the storage 2 to transmit.

The switch 3 transmits the packet (read) received from the control device 4 to the storage 2 based on the instruction (step S1012). The storage 2 returns a response packet (data) to the received packet (read) to the switch (step S1013).

When the switch 3 receives the response packet (data), the switch 3 transfers the response packet (data) to the control device 4 (step S1014).

The control device 4 extracts the entry corresponding to the response packet (data) received from the host-storage correspondence information storage unit 44, and sets the control information (flow entry) in the switch 3 (step S1015). Further, the control device 4 transmits the response packet (data) received in step S1014 to the switch 3 (step S1016), and instructs the host device 1 to transmit it.

The switch 3 transmits the response packet (data) received from the control device 4 to the host device 1 based on the instruction (step S1017). After that, since control information (flow entry) for transferring a packet between the host apparatus 1 and the storage is set in the switch 3, a packet (read) requesting access to the storage 2 transmitted from the host apparatus 1 is read. ) Is directly transferred to the storage 2 (steps S1018 and S1019). Similarly, a response packet (data) transmitted from the storage 2 to the host device 1 is directly transferred to the host device 1 (steps S1020 and S1021).

As described above, according to the present embodiment, when the host device 1 accesses the storage 2, the optimum storage is automatically selected. The reason is that the storage selection device 5 selects the storage based on the service information, and the control device 4 is configured to control the switch 3 based on this.

In the sequence diagram of FIG. 13, each time the switch 3 receives a new packet that does not conform to the existing entry of the flow table 31, the control device 4 transmits the packet to the control device 4, and the control device 4 processes the packet. (Flow entry) is set. For example, as shown in FIG. 14, the control device 4 sets control information (flow entry) for processing the response packet (data) in step S1110. Also good. By doing so, it is possible to omit the transfer of the response packet (data) to the control device 4.

[Second Embodiment]
Next, a second embodiment in which a storage data migration (arrangement) function is added to the first embodiment of the present invention will be described in detail with reference to the drawings. FIG. 15 is a diagram showing a configuration of a storage providing system according to the second embodiment of this invention.

Referring to FIG. 15, the switch 3, the host device 1 and the storages 2a to 2n connected via the switch 3, the control device 4a for controlling the switch 3, and the storage to which the host device 1 is connected are selected. The storage selection device 5a that notifies the control device 4a, the access statistics collection device 7 that collects the access statistical information to the storages 2a to 2n, and the storage management device 8 that performs storage data migration (relocation) processing. A configuration including Hereinafter, a description will be given focusing on differences from the first embodiment.

FIG. 16 is a block diagram showing the configuration of the storage selection device 5a of this embodiment. Referring to FIG. 16, a configuration including a storage characteristic DB 52, an appropriate storage selection unit 51a, and a used storage information storage unit 53 is shown.

The storage characteristic DB 52 is a database that stores information such as the type of each storage, and is the same as the storage characteristic DB 52 of the first embodiment (see FIG. 9).

The used storage information storage unit 53 stores which host device 1 is using which storage. FIG. 17 is a diagram illustrating an example of entries held in a table or the like of the used storage information storage unit 53. Referring to FIG. 17, there is shown an entry in which host information, virtual storage information, and real storage information are associated with each other. In the host information field, the IP address of the host is set. The virtual storage information field is storage information designated by the host device 1 as an access destination, and is set with an IP address (may be a MAC address), an IP protocol number, a TCP / UDP destination port address, and the like. . The actual storage information field is information on storage to be actually accessed, and the storage ID or pointer of the corresponding storage in the storage characteristic DB is set.

The access statistics collection device 7 collects statistical information on access to the storage 2 from the control device 4 and / or the storage 2. Note that the control device 4 can acquire the number of packets that match the rule by acquiring an entry in the flow table 31 of the switch 3 and referring to its Counters field (see FIG. 3). Of course, it is also possible to collect statistical information of direct access from the storage 2. The access statistics information collection timing of the access statistics collection device 7 includes a case where a period defined by the system has passed or a case where an inquiry is received from the storage selection device 5.

The storage management device 8 is a device that moves data between storages. FIG. 18 is a block diagram illustrating a configuration example of the storage management device 8. Referring to FIG. 18, a configuration including a format conversion calculation unit 81 and a data format conversion unit 82 is shown.

The format conversion calculation unit 81 is a logical unit that determines a data conversion method from the storage type before migration, the type of storage after migration, and the data format.

The data format conversion unit 82 is a logical means for converting the data read from the storage before migration into a format to be written to the storage after migration based on the method determined by the format conversion calculation unit 81.

Subsequently, the operation of the present embodiment will be described in detail with reference to the drawings. Since operations of the host device 1, the storage 2, the switch 3, and the control device 4 at the time of packet reception are the same as those in the first embodiment, description thereof is omitted.

FIG. 19 is a flowchart showing the operation of the storage selection device 5a according to the second embodiment of this invention. Referring to FIG. 19, the storage selection device 5a acquires statistical information from the access statistics collection device 7 at a predetermined opportunity (step S401), and extracts application characteristics based on the acquired information (step S402). . Further, the storage selection device 5a selects an appropriate storage from the storage characteristic DB 52 based on the extracted features (step S403). The storage selection based on the extracted features here can use the same method as the storage search (S302 in FIG. 12) in the first embodiment.

If the storage selected in step S403 is the same as the storage recorded in the corresponding entry of the used storage information storage unit 53, the migration process is skipped (to step S401).

On the other hand, when the storage selected in step S403 is not the same as the storage recorded in the corresponding entry of the used storage information storage unit 53 (that is, when a better storage is found), the storage migration process after step S405 is executed. The

First, the storage selection device 5a changes the real storage of the entry of the corresponding host in the used storage information storage unit 53 to the storage selected in Step S403 (Step S405). Next, the storage selection device 5a notifies the control device 4 that the real storage has been changed (step S406), and waits for a response from the control device 4 (step S407). Next, the storage selection device 5a instructs the storage management device 8 to transfer data by designating the migration source storage and the migration destination storage (step S408).

Thereafter, the storage selection device 5a waits for the completion of the data migration (step S409), and upon receiving the data migration completion notification, notifies the control device 4 of the standby release of the migration (step S410). Thereafter, the storage selection device 5a waits until acquisition of the next statistical information (step S411).

Note that the following method can be employed as the storage search method in step S403.
(1) When it is composed of a key and a pair of data, select key value store.
(2) In the case of a table format, a database is selected, and if a column to be searched is designated, an index is created for that column.
(3) When it is composed of a moving image and chapter information, “raw” is selected and the chapter is recorded as an index.

If the file system is specified, if there is additional information that the content is mainly a video, select a file system for large files, and if there is additional information that the text file is the main, If a file system for a small file is selected and there is additional information that the configuration file is the main file system, a file system for a small file and highly available can be selected.

In addition, as the timing of releasing the standby for statistical information acquisition in step S411, when the period specified by the system has passed, or when the instruction from the administrator or the response time of the application exceeds the specified time, There are cases where the response time of the storage exceeds a prescribed time.

FIG. 20 is a flowchart showing the operation of the storage management device 8 according to the second embodiment of this invention. Referring to FIG. 20, first, the storage management device 8 acquires storage information from the storage 2 designated as the migration source and the migration destination (step S701).

Next, the storage management device 8 checks whether or not the data format needs to be changed based on the acquired storage information (step S702). If it is determined that the data format needs to be changed (Yes in step S702), the storage management device 8 derives a data format conversion method at the time of data migration based on the acquired storage information (step S704). ).

When the derivation of the conversion method at the time of data migration is completed or it is determined in step S702 that the data format does not need to be changed, the storage management device 8 initializes the migration destination storage (format, table creation, network setting, etc.) The read target is set to the top (steps S705 and S706).

Subsequently, the storage management device 8 starts data migration processing. First, it is determined whether or not the transfer of all data in the migration source storage has been completed (step S707).

When transfer of all data in the migration source storage has not been completed (No in step S707), the storage management device 8 reads data from the transfer source (step S708). If it is determined in step S702 that the data format needs to be converted (Yes in step S709), the storage management device 8 converts the data format of the data read from the transfer source (step S710).

Next, the storage management device 8 writes the read data subjected to the data format conversion as necessary to the migration destination storage as described above (step S711).

Next, the storage management device 8 advances the position of the data to be read by one, and performs the migration completion determination in step S707. Here, when the transfer of all data in the migration source storage has been completed (Yes in step S707), the storage management device 8 notifies the storage selection device 5a of the completion of migration, and a series of processing ends ( Step S713).

In addition, the following is mentioned as a data conversion method derived | led-out by said step S702.
-When stored in the database, but the search key is only one column, when converting to key value store, the column used for search is stored as key, and the remaining columns are divided between columns The character or length is determined and converted as value.
-Stored as text, but if there are many searches, convert it to a database, disassemble and store the sentence in units of paragraphs, minutes, and phrases, and create an index with the character string used for the search.
-Stored in the key value store, but if the search is often performed again by value, convert it to a database and convert the value to multiple columns.
-When a database is used, but there are many accesses to a specific column, it is converted into a composite database with a memory database, and a column with a high access is placed on the memory.

FIG. 21 is a sequence diagram showing the operation of the storage providing system according to the second embodiment of the present invention achieved by the operation of each device.

Steps S2001 to S2004 show a state in which the host device 1 is accessing the storage 2a selected by the method of the first embodiment or the access statistical information of the present embodiment.

At a predetermined timing, the access statistical information collection device 7 collects access statistical information from the control device 4 and the storage 2a (steps S2005 to S2008).

Thereafter, the storage selection device 5a acquires access statistical information from the access statistical information collection device 7, and determines whether or not the storage device to be accessed by the host device 1 needs to be changed (steps S2009 to 2010). Here, it is assumed that the storage selection device 5a decides to change the storage to be accessed by the host device 1 to the storage 2b and notifies the control device 4 to that effect (step S2011).

Upon receiving the notification, the control device 4 deletes control information (flow entry) that realizes communication between the host 1 and the storage 2a set in the switch 3 (steps S2012 to S2013). Further, the control device 4 instructs the storage management device 8 to transfer the data of the host device 1 in the storage 2a to the storage 2b (step S2014).

Upon receipt of the instruction, the storage management device 8 initializes the storage 2b (steps S2015 to S2016), reads the data in the storage 2a, and starts a write operation to the storage 2b (steps S2017 to S2019). When the migration of the target data is completed, the storage management device 8 notifies the storage selection device 5a of the completion of the data migration (Steps S2020 to S2021).

Upon receipt of the data migration completion notification, the storage selection device 5a notifies the control device 4 of standby release (steps S2022 to S2023).

Thereafter, when the host device 1 sends out a packet (read) requesting access to the storage 2 (step S2024), the switch 3 transfers the packet (read) to the control device 4 (step S2025).

The control device 4 extracts an entry corresponding to the transferred packet (read) from the host-storage correspondence information storage unit 44, and sets control information (flow entry) in the switch 3 (step S2026). Further, the control device 4 transmits the packet (read) transferred in step S2025 to the switch 3 (step S2027), and instructs the storage 2b to transmit.

The switch 3 transmits the packet (read) received from the control device 4 to the storage 2b based on the instruction (step S2028). The storage 2b returns a response packet (data) to the received packet (read) to the switch 3 (step S2029).

When the switch 3 receives the response packet (data), the switch 3 transfers the response packet (data) to the control device 4 (step S2030).

The control device 4 extracts an entry corresponding to the response packet (data) received from the host-storage correspondence information storage unit 44, and sets control information (flow entry) in the switch 3 (step S2031). Further, the control device 4 transmits the response packet (data) received in step S2030 to the switch 3 (step S2032), and instructs the host device 1 to transmit it.

The switch 3 transmits the response packet (data) received from the control device 4 to the host device 1 based on the instruction (step S2033). After that, since control information (flow entry) for transferring a packet between the host device 1 and the storage 2b is set in the switch 3, a packet requesting access to the storage 2b transmitted from the host device 1 ( read) is directly transferred to the storage 2b (steps S2034 and S2035). Similarly, the response packet (data) transmitted from the storage 2 to the host device 1 is directly transferred to the host device 1 (steps S2036 and S2037).

As described above, according to the present embodiment, the host device can access storage more suitable for the operation of the application. At that time, the storage to be accessed can be changed beyond the storage type without changing the application settings or the like.

In this embodiment, instead of inputting service information, storage access history is collected, and the storage selection device selects a more suitable storage based on the collected result. Then, based on the selected result, the storage management device 8 performs data migration. Further, at this time, by converting the data when the storage types are different, the difference in the storage types can be absorbed. Furthermore, since the address grasped by the host device 1 is converted into the address defined in the real storage information by the switch 3, the influence on the application and the like is also hidden.

In the sequence diagram of FIG. 21, each time the switch 3 receives a new packet that does not match the existing entry of the flow table 31, the control device 4 transmits the packet to the control device 4, and the control device 4 processes the packet. (Flow entry) is set. For example, as shown in FIG. 22, the control device 4 sets control information (flow entry) for processing the response packet (data) in step S2126. Also good. By doing so, it is possible to omit the transfer of the response packet (data) to the control device 4.

[Third Embodiment]
Subsequently, a third embodiment in which a protocol conversion function is added to the above-described first embodiment will be described. In the present embodiment, as shown in FIG. 23, the protocol conversion device 9 is arranged in the front stage of the storage 2. Since the other configuration is the same as that of the first embodiment, the difference will be mainly described below.

The protocol conversion device 9 converts the protocol between the host device 1 and the protocol conversion device 9 into a protocol for each storage 2 (for example, a SQL (Structured Query Language) statement or a file I / O). The protocol conversion device 9 is notified in advance of the protocol conversion method for each storage 2 from the storage selection device 5.

In this way, even when various types of storage 2 are connected, the host device 1 can access without knowing the difference.

In the example of FIG. 23, the protocol conversion device 9 is installed outside the storage 2, but the protocol conversion device 9 may be a part of the storage 2.

[Fourth Embodiment]
A configuration in which the protocol conversion function is added to the second embodiment can also be employed. In the example of FIG. 24, the protocol conversion device 9 is arranged in the previous stage of the storage 2 of the second embodiment.

[Fifth Embodiment]
Moreover, the form which united 1st Embodiment and 2nd Embodiment is also employable. In the example of FIG. 25, the appropriate storage selection unit 51a of the storage selection device 5 can perform both storage selection based on service information and storage selection based on access statistical information. Even in such a form, it is possible to select an appropriate storage and perform data migration as necessary.

The preferred embodiments of the present invention have been described above, but the technical scope of the present invention is not limited to the descriptions of the above-described embodiments. For example, in each of the above-described embodiments, the control device 4 and the storage selection device 5 are different devices, but they may be the same device. By doing so, it becomes possible to share the database inside each device and to increase the access speed.

In the second and fourth embodiments, the access statistics collection device 7 is external to the storage selection device 5, but may be the same device. By doing in this way, the communication cost in the case of an analysis is reduced.

Finally, a preferred form of the invention is summarized.
[First embodiment]
(Refer to the storage provision system from the first viewpoint)
[Second form]
In the storage providing system of the first form,
The storage selection device analyzes the characteristics of a service operating on the host device based on access statistics to the storage from each host device, and selects a storage as a storage destination of data used by the host device Offer system.
[Third embodiment]
In the storage providing system of the first or second form,
Furthermore, the storage providing system includes a storage management device that changes a storage that is a storage destination of data used by the host device based on a storage selection result of the storage selection device and instructs the storage to move data .
[Fourth form]
In the third form of storage providing system,
When data used by the host device is moved,
The storage providing system in which the control device changes an instruction to the switch in accordance with movement of data used by the host device.
[Fifth embodiment]
In the storage providing system of the third or fourth form,
The storage management system is a storage providing system that performs data format conversion between the storages that move the data as necessary.
[Sixth embodiment]
In the storage providing system according to any one of the first to fifth aspects,
Furthermore, between the two or more storage units and the host device,
A storage providing system in which a protocol conversion device for converting a message of a first protocol supported by the host device into a message of a second protocol supported by the storage is arranged.
[Seventh form]
In the storage providing system according to any one of the first to sixth aspects,
A storage providing system in which the control device converts a storage address designated by the host device and a storage address actually accessed by the host device.
[Eighth form]
(Refer to the storage selection device from the second viewpoint)
[Ninth Embodiment]
(Refer to the storage provision method from the third viewpoint)
[Tenth embodiment]
(Refer to the program from the fourth viewpoint above.)
Note that the eighth to tenth embodiments can be developed into the second to seventh embodiments as in the first embodiment.

It should be noted that the disclosures of the above patent documents and non-patent documents are incorporated herein by reference. Within the scope of the entire disclosure (including claims) of the present invention, the embodiments and examples can be changed and adjusted based on the basic technical concept. Further, various combinations or selections of various disclosed elements (including each element of each claim, each element of each embodiment or example, each element of each drawing, etc.) within the scope of the claims of the present invention. Is possible. That is, the present invention of course includes various variations and modifications that could be made by those skilled in the art according to the entire disclosure including the claims and the technical idea. In particular, with respect to the numerical ranges described in this document, any numerical value or small range included in the range should be construed as being specifically described even if there is no specific description.

DESCRIPTION OF SYMBOLS 1 Host apparatus 2, 2a-2n Storage 3 Switch 4, 4a Control apparatus 5, 5a Storage selection apparatus 7 Access statistics collection apparatus 8 Storage management apparatus 9, 9a-9n Protocol conversion apparatus 30 Packet processing part 31 Flow table 41 Path | route calculation part 42 Topology database (Topology DB)
43 Flow Entry Setting Unit 44 Host-Storage Corresponding Information Storage Unit 51 Service Analysis Unit 51a Appropriate Storage Selection Unit 52 Storage Characteristic Database (Storage Characteristic DB)
53 Storage Information Storage Unit 81 Format Conversion Calculation Unit 82 Data Format Conversion Unit

Claims (10)

Two or more storages with different characteristics,
A host device that accesses the storage;
A switch disposed between the storage and the host device;
A storage selection device for selecting a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage;
And a control device that controls the switch so that the host device can access the selected storage. The storage selection device analyzes characteristics of a service operating on the host device based on access statistics to the storage from each host device, and selects a storage as a storage destination of data used by the host device. Item 4. The storage providing system according to Item 1. 2. The storage management apparatus according to claim 1, further comprising: a storage management apparatus that changes a storage that is a storage destination of data used by the host apparatus based on a storage selection result of the storage selection apparatus and instructs the storage to move data. Or the storage provision system of 2. When data used by the host device is moved,
The storage providing system according to claim 3, wherein the control device changes an instruction to the switch according to movement of data used by the host device. The storage providing system according to claim 3 or 4, wherein the storage management device performs data format conversion between the storages that move the data, if necessary. Furthermore, between the two or more storage units and the host device,
The storage providing system according to any one of claims 1 to 5, wherein a protocol conversion device that converts a message of a first protocol supported by the host device and a message of a second protocol supported by the storage is disposed. The storage control system according to any one of claims 1 to 6, wherein the control device performs conversion between a storage address designated by the host device and a storage address actually accessed by the host device. Two or more storages with different characteristics,
A host device that accesses the storage;
A switch disposed between the storage and the host device;
A control device for controlling the switch, and a storage selection device connected to the control device,
A storage selection unit that selects a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage;
A storage selection device comprising: a notification unit that notifies the control device of a combination of the selected storage and a host device, and controls the switch so that the host device can access the selected storage. Two or more storages with different characteristics,
A host device that accesses the storage;
A switch disposed between the storage and the host device;
In a storage providing system to which a control device for controlling the switch is connected,
Selecting a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage;
A storage providing method comprising: notifying the control device of a combination of the selected storage and a host device, and controlling the switch so that the host device can access the selected storage. Two or more storages with different characteristics,
A host device that accesses the storage;
A switch disposed between the storage and the host device;
A computer connected to a storage providing system to which the control device for controlling the switch is connected;
Processing for selecting a storage to be used by the host device based on characteristics of a service operating on the host device and characteristics of the storage;
A program for notifying the control device of a combination of the selected storage and host device, and executing processing for controlling the switch so that the host device can access the selected storage.

Images