JP4622914B2 - Replication volume creation device, replication volume creation method, replication volume creation program - Google Patents

Description

  The present invention relates to a replica volume creation technique, and more particularly, to a replica volume creation method that improves capacity efficiency and has physical failure resistance.

As a method of creating a replica of a volume in a conventional disk array device, a method of securing a volume of the same capacity and acquiring a complete copy (clone, mirroring) of the volume, and saving only the update difference and saving with the master volume There is a known method (snapshot) for securing a copy from the updated difference.
In general, mirroring obtains a complete replica of a volume, so that the replica data is not lost even if the master volume becomes a physical failure. However, even if there is an unused area in the master volume, a volume with the same capacity as the master volume is required, so capacity efficiency is poor.
On the other hand, since the snapshot saves only the update difference, the capacity efficiency is good. However, since data in the unupdated portion is read from the master volume, duplicate data is lost when the master volume becomes a physical failure.

Patent Document 1 describes an example of a method for creating a replica of a volume. In this method, when the snapshot is not taken, the write data is duplicated to reduce the load on the CPU of the computer, the traffic on the data communication path, and the load on the external storage device.
Patent Document 2 describes an example of a method for improving capacity efficiency when a multi-generation copy is created by combining mirroring and snapshot.
Patent Document 3 describes a capacity virtualization method using an FC (Fiber Channel) switch. In this method, when a virtual volume is created, a real area is not allocated to the virtual volume, but a real area is allocated when the host processor writes to the virtual volume, thereby efficiently allocating a storage area. .

JP 2001-306407 A JP-A-2005-250676 JP 2005-011316 A

The method described in Patent Document 1 has a problem that capacity efficiency cannot be improved.
In the method described in Patent Document 2, capacity efficiency can be improved, but there is a problem that there is no effect unless a plurality of generations of replicas are configured.
The capacity virtualization method described in Patent Document 3 has a problem that it cannot be applied to creating a replica in a disk array device. In addition, if there is a shortage of capacity in the actual usage area for the virtual volume, there is a possibility that I / O to the master volume may be stopped, so there is a problem that it cannot be used in a system that holds important data.

  The present invention provides a method for creating a replica in a disk array device that solves capacity efficiency and physical failure resistance of a master volume, regardless of the number of generations and without affecting the master volume.

The replication volume creation apparatus of the present invention is an input / output processing means for executing data input processing and output processing for a master volume in accordance with an input / output command from the host computer, and replication of the master volume in accordance with the replication command from the host computer. Replica creation means for creating a replica volume, and actual use area information for storing information indicating whether each area of the master volume is an actual use area storing data or an unused area not storing data A storage table is provided.
The input / output processing means has a function of updating the information corresponding to the area of the actual usage area information storage table so as to indicate that the area is the actual usage area when the area to be input processed is an unused area. I have.
The copy creation means refers to the actual use area information storage table and creates a copy volume by writing only the data stored in the actual use area to the copy data storage unit. And a function of creating a copy of the used area information storage table.

  According to the duplicate volume creation device, the actual processing area information storage table stores whether each area of the master volume is an actual usage area or an unused area. Only the usage area can be replicated in the replicated data storage. Therefore, the capacity of the replication volume can be saved.

In the duplicate volume creation device, the duplicate creation means has a function of associating and storing the actual use area and the area of the duplicate data storage unit in which the data stored in this area is duplicated in the duplicate data information storage table. (Claim 2).
In this way, the correspondence relationship between the actual usage area of the master volume and the area of the replicated data storage unit in which the data stored in this area is replicated is stored in the replicated data information storage table. Even if there is a physical failure, the replica data of the master volume can be maintained.
In addition, since only the replication volume is virtually managed, even if the actual capacity of the replication data storage unit is insufficient, the master volume is not affected and a robust system can be constructed.

In the duplicate volume creation device, the duplicate creation means may create a new duplicate data information storage table each time a duplicate volume is created.
In this way, the replicated data can be efficiently copied from the master volume to the replicated data storage unit.

In the replication volume creation device, the input / output processing means determines whether or not the target data to be input / output is stored in the replication volume when receiving an input / output command for the replication volume from the host computer. Judgment is made with reference to the storage table, and if it is not stored, the replication data storage area where the target data is stored is stored in the replication data information storage table, and the actual usage area information storage table is targeted for replication You may make it provide the function to update so that the area | region where data is stored may be an actual utilization area | claim (Claim 4).
In this way, even if there is an input / output to the replicated data storage unit for purposes other than creating a replica, the replica of the actual usage area information storage table after the input / output and the replicated data information storage table should be maintained consistently. I can do it.

  In the replica volume creation method of the present invention, when an input command is received from a host computer to an unused area that does not store master volume data, the input / output processing means stores data for each area of the master volume. Updates the information corresponding to the area in the actual used area information storage table that stores information indicating whether it is an actual used area or an unused area and executes an input command. When a copy command is received from the master computer and the input processing step, the copy creation means refers to the actual use area information storage table and writes only the data stored in the actual use area to the copy data storage unit. To create a replica volume of the master volume and store the actual usage area information storage table at the time when the replication command was issued. By the replica creation process for creating manufacturing to create a duplicate volume (claims 5 to 8).

  According to the duplicate volume creation method described above, whether each area of the master volume is an actual used area or an unused area is stored in the actual used area information storage table. Only the usage area can be replicated in the replicated data storage. Therefore, the capacity of the replication volume can be saved.

In the duplicate volume creation method, in the duplicate creation step, the actual use area and the duplicate data storage area where the data stored in this area is duplicated are stored in the duplicate data information storage table in association with each other. (Claim 6).
In this way, the correspondence relationship between the actual usage area of the master volume and the area of the replicated data storage unit in which the data stored in this area is replicated is stored in the replicated data information storage table. Even if there is a physical failure, the replica data of the master volume can be maintained.
In addition, since only the replication volume is virtually managed, even if the actual capacity of the replication data storage unit is insufficient, the master volume is not affected and a robust system can be constructed.

In the duplicate volume creation method, a duplicate data information storage table may be newly created each time a duplicate volume is created in the duplicate creation step.
In this way, the replicated data can be efficiently copied from the master volume to the replicated data storage unit.

In the above replica volume creation method, when the host computer receives an input / output command for the replica volume after completion of the replica creation process, the replica data indicates whether or not the target data to be input / output is stored in the replica volume. If it is determined by referring to the information storage table and is not stored, the area of the duplicate data storage unit where the target data is stored is stored in the duplicate data information storage table, and the actual use area information storage table is duplicated. You may make it update so that the area | region where object data is memorize | stored may show that it is an actual utilization area | region (Claim 8).
In this way, even if there is an input / output to the replicated data storage unit for purposes other than creating a replica, the replica of the actual usage area information storage table after the input / output and the replicated data information storage table should be maintained consistently. I can do it.

  The replica volume creation program of the present invention stores the data for each area of the master volume when the computer receives an input command for an unused area that does not store the data of the master volume from the host computer. An input process for updating the information corresponding to the area in the actual use area information storage table that stores information indicating whether the area is an unused area or not, and executing an input command When a replication command is received from the function and the master computer, a replication volume of the master volume is created by referring to the actual usage area information storage table and writing only the data stored in the actual usage area to the replication data storage unit At the same time, create a copy of the actual used area information storage table at the time when there was a copy command. Creating a duplicate volume by executing the replica creation function (claims 9 to 12).

  According to the above-described duplicate volume creation program, whether each area of the master volume is an actual used area or an unused area is stored in the actual used area information storage table. Only the usage area can be replicated in the replicated data storage. Therefore, the capacity of the replication volume can be saved.

In the above replica volume creation program, the replica creation function may store the actual use area in the duplicate data information storage table in association with the duplicate data storage area in which the data stored in this area is duplicated. Good (claim 10).
In this way, the correspondence relationship between the actual usage area of the master volume and the area of the replicated data storage unit in which the data stored in this area is replicated is stored in the replicated data information storage table. Even if there is a physical failure, the replica data of the master volume can be maintained.
In addition, since only the replication volume is virtually managed, even if the actual capacity of the replication data storage unit is insufficient, the master volume is not affected and a robust system can be constructed.

In the duplicate volume creation program, the duplicate creation function may create a new duplicate data information storage table each time a duplicate volume is created.
In this way, the replicated data can be efficiently copied from the master volume to the replicated data storage unit.

In the above replica volume creation program, when the host computer receives an I / O command to the replica volume after the replica volume creation by the replica creation function is completed, the target data to be input / output is stored in the replica volume on the computer. If the data is not stored, the area of the duplicate data storage unit where the target data is stored is stored in the duplicate data information storage table. A function of updating the use area information storage table so as to indicate that the area in which the target data is stored is the actual use area may be executed (claim 12).
In this way, even if there is an input / output to the replicated data storage unit for purposes other than creating a replica, the replica of the actual usage area information storage table after the input / output and the replicated data information storage table should be maintained consistently. I can do it.

  According to the present invention, since each area of the master volume is an actual use area or an unused area is stored in the actual use area information storage table, the replication processing means refers to this, and only the actual use area Can be replicated in the replicated data storage. Therefore, the capacity of the replication volume can be saved.

Hereinafter, the configuration and operation of the replication volume creation system 1 according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a functional block diagram of a disk array device 10 (an example of a duplicate volume creation device).
Referring to FIG. 1, a replication volume creation system 1 is, for example, a SAN (Storage Area).
A disk array device 10 and a computer 11 (an example of a host computer) that are connected by a network and can communicate with each other.
The disk array device 10 includes an input / output processing unit 100, a copy creation unit 101, a master volume 110, a copy data storage unit 111, an actual use area information storage table 120, a copy 121 of an actual use area information storage table, and copy data information. A storage table 122 is included. The master volume 110, the replicated data storage unit 111, the actual usage area information storage table 120, the actual usage area information storage table replica 121, and the replication data information storage table 122 are stored in an arbitrary storage device such as a hard disk device or a nonvolatile memory. Stored.
The computer 11 operates under program control and issues a data input / output request to the master volume 110 and a replication volume creation request to the disk array device 11.

Each of these means generally operates as follows.
The disk array device 10 processes an input / output request to the master volume 110 from the computer 11 connected to the disk array device 10 by the input / output processing means 100. At this time, if an input / output occurs in an unused area of the master volume 110, the input / output processing unit 100 updates the actual used area information storage table 120 to obtain information on the actual used area for the master volume 110. Update.

  When there is a replication volume creation request, the replication creation unit 101 copies the actual usage area information storage table 120 at the time of the creation request to the replication 121 of the actual usage area information storage table. Further, the actual use area of the master volume 110 at the time when the creation request is made is copied to the copy data storage unit 111 with reference to the actual use area information storage table 120. Copy information to the copy data storage unit 111 is reflected in the copy data information storage table 122.

  As shown in FIG. 2, the “real use area” refers to the area 110a in which the data 310 is stored in the storage area of the master volume 110, and the “unused area” refers to the storage of the data 310. It refers to the region 110b that is not.

  The input / output processing unit 100 and the replica creation unit 101 are realized by a CPU included in the disk array device 10 executing a program and controlling the hardware of the disk array device 10.

Next, the operation of the replication volume creation system 1 will be described in detail with reference to FIGS. 1, 2, 3, 4, 5, 6, and 7.
First, the operation of the duplicate volume creation system 1 when the computer 11 of FIG. 1 uses the master volume 110 of the disk array device 10 will be described in detail.
Referring to the flowchart of FIG. 3, when an input / output request (step A <b> 1 in FIG. 3) is made to the master volume 110 of the disk array device 10 by the computer 11 of FIG. 1, the input / output request destination is the master volume 110. The input / output processing means 100 updates the actual usage area information so that the input / output request destination is the actual usage area in the actual usage area information storage table 120 ( Step A2 in FIG. 3), the requested processing is performed (step A4 in FIG. 3). Then, a response of processing completion is sent to the computer 11 (step A4 in FIG. 3).

  For example, as shown in FIG. 4, the actual use area information storage table 120 divides a volume into blocks and manages the blocks in real use areas and unused areas. In FIG. 4, the actual use area is hatched (for example, 0c block, reference numeral 120a) and the unused area is blank (for example, 12 block, reference numeral 120b) so that it can be easily understood visually. Actually, it is assumed that the array data corresponds to one element for each block, the actual use area has a value “1”, and the unused area has a value “0”.

  When the input / output request destination is the actual use area of the master volume 110, the input / output processing means 100 processes the input / output request to the master volume 110 without updating the information of the actual use area. (Step A3 in FIG. 3), and sends a process completion response to the computer 11 (Step A4 in FIG. 3).

Next, the operation of the replication volume creation system 1 when a replication volume creation instruction is issued to the master volume 110 in FIG. 1 will be described in detail.
Referring to the flowchart of FIG. 5, when there is a replication volume creation request to the master volume 110 (step B1 in FIG. 5), the replication creation unit 101 determines the actual usage area information at the time of the creation request. The storage table 120 is copied to the copy 121 of the actual usage area information storage table (step B2 in FIG. 5).
If the data copy of the actual use area of the master volume 110 is not completed to the duplicate data storage unit 111, the information on the area to be copied to the empty area of the duplicate data storage unit 111 is stored in the duplicate data information storage table 122. The data is reflected (step B3 in FIG. 5), and the data in the master volume 110 at the time of the creation request is copied to the duplicate data storage unit 111 (step B4 in FIG. 5). At this time, a new management table may be provided in order to efficiently copy the duplicate data.

For example, as shown in FIG. 6B, the duplicate data information storage table 122 is obtained by dividing the duplicate data storage unit 111 into blocks and copying the actual use area information storage table to the copy source master volume area. The table number 121 and the block address are stored. For example, the data indicated by reference numeral 122a in FIG. 6B is stored in the “10” block of the duplicate data storage unit in the “16” block of the replica 121 of the actual usage area information storage table whose table number is “1”. It shows that corresponding data is stored.
When the data copy of the actual use area of the master volume 110 at the time of the creation request to the duplicate data storage unit 111 is completed (step B5 in FIG. 5), the creation of the duplicate is completed.

Next, the operation of the replica volume generation system 1 when the computer 11 uses the replica volume after creating the replica of the master volume 110 will be described in detail. The actual usage area information storage table replica 121 and the replicated data information storage table 122 in the description use the example of FIG.
As shown in the flowchart of FIG. 7, when the computer 11 makes an input / output request (step C1 in FIG. 7) for the master volume 110 replication, the requested data is stored in the replicated data storage unit 111. Is searched from the duplicate data storage information table 122 (step C2 in FIG. 7). In FIG. 6, since the address of the actual use area for the copy exists in the copy 121 of the actual use area information storage table and these addresses are mapped in the copy data storage information table 122, the data for which an input / output request has been made Can be retrieved from the duplicate data storage information table 122 as to whether or not is stored in the duplicate data storage unit 111.
If the requested data is stored in the duplicate data storage unit 111, the duplicate data storage unit 111 is processed for the input / output request for that area (step C5 in FIG. 7).

If the requested data is not stored in the duplicate data storage unit 111, the replica 121 of the actual use area information storage table is updated (step C3 in FIG. 7), and the address information of the actual use area for the duplicate is updated. To do. Information about the area where new input / output is performed with respect to the empty area of the duplicate data storage unit 111 is reflected in the duplicate data information storage table 122 (step C4 in FIG. 7), and the new duplicate data storage unit 111 area is reflected. On the other hand, the input / output request is processed (step C5 in FIG. 7).
For example, when there is an input command for the area 11 of the duplicate data storage unit 111, the operation is as follows. First, the input / output processing unit 100 refers to the duplicate data information storage table 122 and determines that the data to be written is not stored in the duplicate data storage unit 111 because “11” is not found. Next, the location (symbol 121a in FIG. 6A) corresponding to the area 11 of the replica 121 of the actual usage area information storage table is updated to indicate that it is an actual usage area. Then, the location of the area 81 in the duplicate data information storage table 122 (reference numeral 122b in FIG. 6B) is updated to “1-11”, and the designated data is written in the area 81 of the duplicate data storage unit.
When the input / output processing is completed, the fact is reported to the computer 11 (step C6 in FIG. 7).

In the present invention, the replicated data is managed by the actual use area information storage table, the duplicate of the actual use area information storage table, and the duplicate data information storage table. It can also be set as the structure provided with the some replication data storage part.
For example, as shown in FIG. 8, a replica can be created for one replica data storage unit 1112 from a plurality of master volumes of a first master volume 1110 and a second master volume 1111.
Further, as shown in FIG. 9, the master volume replica data is distributed to a plurality of replica data storage units of the first replica data storage unit 2111 and the second replica data storage unit 2112 for the master volume 2110. You can also.

Next, effects of the duplicate volume creation system 1 will be described.
The first effect is that the capacity of the replication volume can be saved.
The reason is that the actual used area of the master volume 110 is managed by the actual used area information storage table 120, and the replica creating unit 101 does not secure the unused area in which data is not secured as a replica, This is because only the actual use area in which 110 data is secured is reserved as a copy area.
The second effect is that even if a physical failure occurs in the master volume 110, the duplicate volume is not lost.
The reason is that all the actual use areas in which the master volume data is secured are copied to the replica, and the correspondence between the actual use area and the copy data storage area in which the copy is stored is stored in the copy data information storage table 122. It is because it remembers.

1 is a block diagram of a replication volume creation system that is an embodiment of the present invention. FIG. It is a figure explaining the definition of an unused area | region and an actual utilization area | region. It is a flowchart showing the operation of the replication volume creation system when there is an I / O request to the master volume. It is a figure which shows the example of a data structure of an actual utilization area | region information storage table. It is a flowchart showing the operation of the replication volume creation system when there is a replication volume creation request, FIG. 6A is a diagram illustrating an example of a data structure of a copy of the actual usage area information storage table. FIG. 6B is a diagram showing an example of the data structure of the duplicate data information storage table. It is a flowchart showing the operation of the replication volume creation system when there is an input / output request to the replication volume. It is a figure which shows the modification of the replication volume production system of FIG. It is a figure which shows the modification of the replication volume production system of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Duplicated volume creation system 10 Disk array apparatus 11 Computer 100 Input / output processing means 101 Duplicate creation means 110 Master volume 111 Duplicated data storage part 120 Actual use area information storage table 121 Duplication of actual use area information storage table 122 Duplicated data information storage table

Claims (12)

Input / output processing means for executing data input processing and output processing for the master volume in accordance with an input / output command from the host computer, and replica creation means for creating a replica volume that is a replica of the master volume in accordance with the replication command from the host computer In the replica volume creation device with
An actual use area information storage table for storing information indicating whether each area of the master volume is an actual use area storing data or an unused area not storing data;
The input / output processing means indicates that the information corresponding to the area of the actual use area information storage table is the actual use area when the area to be subjected to the input process is the unused area. With the ability to update
The copy creating means refers to the actual use area information storage table, and creates the copy volume by writing only the data stored in the actual use area to the copy data storage unit; A duplicate volume creation device comprising a function of creating a duplicate of the actual use area information storage table at a certain point in time.   The replica creating means has a function of associating and storing the actual use area and the area of the replicated data storage unit in which the data stored in the area is replicated in a replicated data information storage table The replication volume creation device according to claim 1.   3. The duplicate volume creation device according to claim 2, wherein the duplicate creation unit newly creates the duplicate data information storage table every time the duplicate volume is created.   When the input / output processing means receives an input / output command for the replication volume from the host computer, the replication data information storage table indicates whether or not the target data to be input / output is stored in the replication volume. If the target data is stored in the duplicate data information storage table, the duplicate data storage area is stored, and the actual use area information storage table is duplicated. The replication volume creation device according to claim 2 or 3, further comprising a function of updating the area in which the target data is stored to indicate that the area is the actual use area. When an input command is received from the host computer to an unused area that does not store master volume data, the input / output processing means is an actual used area that stores data for each area of the master volume. An input processing step of updating the information corresponding to the area in the actual use area information storage table storing information indicating whether the area is an unused area to indicate that the area is the actual use area and executing the input command; ,
When receiving a copy command from the master computer, the copy creating means refers to the actual use area information storage table and writes only the data stored in the actual use area to the copy data storage unit. A replication volume creation method comprising: creating a replication volume of a volume and creating a replication of the actual use area information storage table at the time when the replication command is issued.   The duplicate creation step is characterized in that the actual use area and the area of the duplicate data storage unit in which the data stored in the area is duplicated are stored in a duplicate data information storage table in association with each other. 5. A replication volume creation method according to 5.   7. The duplicate volume creation method according to claim 6, wherein in the duplicate creation step, the duplicate data information storage table is newly created every time the duplicate volume is created.   When the host computer receives an input / output command for the replica volume after the replica creation step, the replica data information storage stores whether or not the target data to be input / output is stored in the replica volume. If it is determined by referring to the table and is not stored, the area of the duplicate data storage unit where the target data is stored is stored in the duplicate data information storage table, and the actual use area information storage table 8. The replication volume creation method according to claim 6, wherein the replication is updated so as to indicate that the area in which the target data is stored is the actual usage area. On the computer,
When an input command is received from a host computer for an unused area that does not store master volume data, whether it is an actual used area that stores data for each area of the master volume or the unused area An input processing function for updating the information corresponding to the area of the actual use area information storage table that stores information indicating the actual use area and executing the input command;
When a replication command is received from the master computer, the replication volume of the master volume is created by referring to the actual usage area information storage table and writing only the data stored in the actual usage area to the replication data storage unit. A replication volume creation program that creates and executes a replication creation function for creating a copy of the actual use area information storage table at the time when the replication command is issued.   The duplicate creation function is characterized in that the actual use area and the area of the duplicate data storage unit in which the data stored in this area is duplicated are associated and stored in the duplicate data information storage table. The duplicate volume creation program according to claim 9.   11. The duplicate volume creation program according to claim 10, wherein the duplicate creation function newly creates the duplicate data information storage table every time the duplicate volume is created.   After the creation of the replication volume by the replication creation function, when receiving an input / output command for the replication volume from the host computer, the target data to be input / output is stored in the replication volume in the computer. Whether the target data is stored in the replicated data information storage table if not stored. And executing a function of updating a copy of the actual use area information storage table so as to indicate that the area in which the target data is stored is the actual use area. 11. A replication volume creation program according to 11.

Images