JP2008158570A - Storage system, replication method for logic volume and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a storage system or the like, holding down the activity ratio of a communication line connecting the storage devices in performing replication. <P>SOLUTION: A primary storage device and a secondary device each include a check code creating means for reading data in a comparison object from a master volume or a duplicate volute of own devices, creating a check code for the data and storing the same in a memory. One of the primary storage device and the secondary storage device includes a data comparing means for comparing a check code of the own device with a check code of the other device obtained through the communication line, and informing the comparison result to the other, and the primary storage device and the secondary storage device respectively include differential map updating means for updating a differential map corresponding to the comparison object range to show that there is no difference between the master volume and the duplicate volume when a check code matching occurs between the own device and the other device. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a logical volume replication technique, and more particularly to a storage system or the like that can keep the usage rate of a communication line connecting storage apparatuses low when performing replication.

  The replication function provided in a storage device such as a disk array device manages the update IO received from the host after constructing a master volume and replica volume pair as a differential map. When a synchronization instruction is received from the host, a differential copy is made according to the differential map. It is generally known to do this as described in Patent Document 1.

  According to the difference map, only the area updated from the host is copied and there is an advantage that a full difference copy is not made each time.

JP 2000-330730 A

  However, in the conventional technology, copying is performed according to the difference map even in the following cases, so that the synchronization of the pair is established when the device having the master volume and the device having the replication volume are in remote locations and the line bandwidth connecting the devices is small. There is a problem that it takes time to obtain the copy, and originally unnecessary copy data flows through the line.

(First case)
Immediately after the construction of the logical disk, the master volume and the replication volume are both initial data and are likely to have the same data. However, depending on the host computer to which the disk array device is connected, a slight difference in data may occur by writing the volume label.
Even in such a case, since the identity of the data cannot be guaranteed immediately after the pair volume is constructed, registration to the difference map corresponding to all data storage areas is performed, and the first copy is a full difference copy that copies all areas. Become.

(Second case)
When the master volume and the replica volume are synchronized and then separated, and the replica volume is backed up and the computer system is tested using the replica volume, the replica volume backup is copied to the replica volume after the test is completed. .
At this point, the master volume and the replication volume are synchronized, but since the difference map is updated by the update IO from the host computer to the replication volume, a full differential copy is obtained.

  Accordingly, an object of the present invention is to provide a storage system or the like that can keep the usage rate of a communication line connecting storage devices low when performing replication.

The first storage system of the present invention includes a primary storage device and a secondary storage device, and each of the primary storage device and the secondary storage device is designated by the host device when the primary storage device receives a verification instruction from the host device. The comparison target range data is read from the master volume or replication volume of the own device, and a check code creating means for creating a check code for this data and storing it in the memory is provided, one of the primary storage device and the secondary storage device is And a data comparison means for comparing the check code read from the memory of its own device with the check code of the other device acquired via the communication line and notifying the comparison result to the other, and comprising a primary storage device and a secondary storage device Respectively,
Difference map update means is provided for updating the difference map corresponding to the comparison target range to indicate that there is no difference between the master volume and the replication volume when the check codes of the own device and the other device match.

According to the above storage device, when performing replication between the master volume and the replicated volume, it is recorded in the comparison range of the volume that transmission / reception is required via the communication line connecting the primary storage device and the secondary storage device. It is only the check code of the data.
Therefore, it is possible to keep the usage rate of the communication line connecting the storage devices low when performing replication.

  The second storage system of the present invention comprises a primary storage device and a secondary storage device, and one of the primary storage device and the secondary storage device has its own device when the primary storage device receives a verification instruction from the host device. The primary storage device and the secondary storage are provided with data comparison means for comparing the data read from the file update table storage area with the data in the file update table storage area of another device acquired via the communication line and notifying the comparison result to the other. Each device shows a difference map corresponding to the verified range specified by the host device when the data in the file update table storage area of its own device and that of the other device match, indicating that there is no difference between the master volume and the replication volume. Day to update And it includes a comparison means.

According to the above storage device, when replicating the master volume and the replication volume, it is necessary to send and receive via the communication line connecting the primary storage device and the secondary storage device in the storage area of the file update table. It is only the data that is being processed.
Therefore, it is possible to keep the usage rate of the communication line connecting the storage devices low when performing replication.

  In the first or second storage system, the check code comparison unit, the data comparison unit, and the difference map update unit may operate immediately after the pair of the master volume and the replication volume is constructed. 3).

  In the first or second storage system, the check code comparison unit, the data comparison unit, and the difference map update unit may operate immediately after the replication volume is restored by backup of the replication volume. ).

  The logical volume replication method and program according to the present invention can also solve the above-mentioned problems in the same manner as the above storage system.

The first effect is that the time required for replication can be saved.
According to the present invention, the data comparison means verifies the data necessary for verifying the identity of the master volume and the replica volume, and if this data is the same, by deleting the difference map, This is to avoid copying unnecessary data.
The second effect is that the line cost for relaying between disk arrays can be saved.
The data sent and received between the disk arrays via the communication line is only the data necessary for verifying the identity, and the data that matches between the master and replica volume that was flowing through the communication line in the replication using the conventional method This is because it is no longer necessary to flow.

The configuration and operation of the storage system 1 according to an embodiment of the present invention will be described with reference to FIGS. 1, 2, and 3. FIG.
FIG. 1 is a diagram showing the configuration of the storage system 1. The storage system 1 includes a primary host 10, a secondary host 20, a disk array device 11, and a disk array device 21. The disk array device 11 and the disk array device 21 are connected to the primary host 10 and the secondary host 20, respectively, and receive and process read / write IOs and synchronization instruction commands. The primary host 10 and the disk array device 112 constitute a primary site 12, and the secondary host 10 and the disk array device 21 constitute a secondary site 22.
The primary site 12 and the secondary site 22 are installed at geographically remote locations, but the disk array device 11 and the disk array device 21 are communicably connected via the network 2 and will be described later during normal operation. The master volume 120 and the replication volume 220 are replicated using the difference map created by the difference map creation circuit 100 (200).

  The disk array device 11 (21) includes a difference map creation circuit 100 (200), a check code creation circuit 101 (201), a data comparison circuit 102 (202), a memory 110 (210), and a master volume 120 (replication volume 220). ).

The master volume 120 is a logical volume that is recognized by the primary host 10 as a single storage area, and is physically composed of one or a plurality of disk devices. The duplicate volume 220 is a logical volume that duplicates the master volume 120. The synchronization (replication) of the master volume 120 and the replication volume 220 is performed in response to a synchronization instruction from the primary host 10. At that time, only the updated area is copied with reference to the difference map 111.
N pairs of the master volume 120 and the replica volume 220 are constructed as indicated by reference numerals MV1 to MVn (RV1 to RVn) in FIG. The following description will be given focusing on any one of these.

  The memory 110 (210) is configured by, for example, a RAM (Random Access Memory), and stores a difference map 111 (211) and a check code 112 (21) of the master volume 120 (replicated volume 220).

  The difference map creation circuit 100 (200) creates a difference map 111 (211) when processing an update IO for the master volume 120 (replication volume 220) and stores it in the memory 110 (210).

  The check code creation circuit 101 (201) reads data stored in the comparison range designated from the primary host 10 from the master volume 120 (replication volume 220), creates a check code for the data, and 110 (210). The check code is, for example, a hash value of the read data, and is created so that the size is smaller than the original data.

The data comparison circuit 102 (202) has the following three functions.
(1) The check data read from the memory 110 (210) of the own device is compared with the check code of the other device acquired via the network 2, and the comparison result is notified to the other device (function as data comparison means). .
(2) The data of the file update table area read from the master volume 120 (replica volume 220) of the own apparatus is compared with the data of the file update table area of the other apparatus acquired via the network 2, and the comparison result is sent to the other apparatus. Notify (function as data comparison means).
(3) As a result of the comparison in (1) or (2) above, if the check code or the file update table area data match between the own device and another device, it is designated from the primary host 10 in the difference map The portion corresponding to the range is deleted (value is set to 0), and the difference map 111 (211) is updated so as to indicate that there is no difference between the master volume 120 and the replication volume 220 (as a difference map update means) Function of).
Of the above three functions, (3) is necessary for both the data comparison circuit 102 and the data comparison circuit 202, but only one of the functions (1) and (2) may be provided. Good. Hereinafter, only the data comparison circuit 102 has the functions (1) and (2), that is, the disk array device 11 compares the data in the check code or the file update table area, and notifies the disk array device 21 of the comparison result. It will be described as being.

  FIGS. 2 and 3 are diagrams showing how the data 113 and 213 stored in the master volume 120 and the replication volume 220 and the corresponding difference maps 111 and 211 change depending on the operation of the storage system 1.

First, a pair relationship in which a logical volume in the disk array device 11 and a logical volume in the disk array device 21 become a master volume 120 and a replication volume 220 is constructed by a pair construction instruction command from the primary host 10. This is called the first state, and an operation explanatory diagram from this state is shown in FIG.
In this state, the identity of the data between the two volumes cannot be guaranteed. Therefore, in a general method, a full differential copy from the master volume 120 to the duplicate volume 220 is performed, and the pair is once synchronized.

  During operation, the pair is used in the synchronized state or in the separated state based on the policy. However, even when operating in the synchronized state, the pair is in the synchronized state when performing a secondary site operation test using a replicated volume. Separated from This is called the second state. FIG. 3 shows an operation explanatory diagram from this state.

In the first state, before the pair construction, the data that becomes the master volume 120 and the data that becomes the replication volume 220 are completely the same or slightly different, but the data storage area is clear. Assume that
This state is represented by the state 500 in FIG. 2, and even though the data 113 and the data 213 match, the difference maps 111 and 211 are set to “1” bits indicating that there is a difference. It will be in the state.
When the disk array device 11 receives a command for verifying the identity of the master volume 120 and the replicated volume 220 from the primary host 10, the disk array device 11 communicates with the disk array device 21, and the disk array devices 11, 21 communicate with each other. Start verification of data identity.

There are two methods for verifying data identity, and they can be used properly according to the application.
The first method is used when a file system in which a file update table in a logical volume is stored is difficult to specify, and is a data storage area (comparison target area) instructed by the primary host 10. Are read out from the logical volumes 120 and 220, check codes are created through the check code creation circuits 101 and 201, and stored in the memories 110 and 210.
The data comparison circuit 102 reads the check code of at most several bytes created in this way from the memory 110 for the disk array device 11 and acquires it from the disk array device 21 via the network 2 for the disk array device 21. Are compared, and the comparison result is notified to the data comparison circuit 202.

  The second method is a method used in the case of a file system in which the area in which the file update table in the logical volume is stored can be specified. The data in the file update table area is transferred to the logical volume 120 for the disk array device 11. The disk array device 21 is obtained from the disk array device 21 via the network 2 and compared, and the comparison result is notified to the data comparison circuit 202 (state 510 in FIG. 2).

In both cases of the first method and the second method, if the comparison data match, the memory is assumed to have verified the identity of the data with respect to the identity data verification range instructed from the primary host 10. The difference map corresponding to the verified area stored in 110 and 210 is deleted. This state is represented by the state 520 in FIG. 2, and the difference maps 111 and 211 are in a state where a bit of “0” indicating that there is no difference is set at a location where the identity data has been verified.
The identity data verification range coincides with the above comparison control range in the first method. In the second method, if a specific block in which the file update tables of both the master and duplicate volumes are recorded, for example, the first block matches, it can be determined that the entire volume matches. The property data verification range is the entire volume (an example of a verified range).

  In the second state, it is assumed that the secondary host 20 performs a system test using the duplicate volume 220 that is the same data as the primary site. At this time, it is assumed that the backup 230 of the duplicate volume is prepared in advance (state 800 in FIG. 3).

In the update IO from the secondary host 20 at the time of the system test, data is generally written to the replication volume by the disk array device 21 and the update location information of the data is registered in the difference map 211 of the memory 210 through the difference map creation circuit 200. Is done.
This state is represented by the state 810 in FIG. 2, and the difference map 211 is in a state in which a bit “1” indicating that there is a difference is set at the update IO position.

When the system test is completed, the data before the test is restored using the backup 230 of the duplicate volume.
At this time, the same data as the master volume 120 of the primary site 12 is obtained, but the duplicate volume 220 holds a state registered in the difference map as an updated IO difference (state 820 in FIG. 3).

In this state, the identity of the data is verified by the first method or the second method (state 830 in FIG. 3), and the difference map 720 is deleted. This state is represented by the state 840 in FIG. 3, and the difference map 720 is in a state where a bit of “0” indicating that there is no difference is set at a location where the identity data has been verified.

In the above embodiment, the master volume and the replication volume provided in the disk array device are described as an example of a one-to-one pair relationship in a remote location. However, if the master volume and the replication volume are on the same campus, The present invention is also applicable when there are two or more replica volumes for one master volume.
Further, the present invention can be applied not only to the disk array device but also to other types of storage devices.
Further, the same operation as that performed by the difference map creation circuit 100 (200), the check code creation circuit 101 (201), and the data comparison circuit 102 (202) may be realized by a computer program executed by the disk array device 11 (21). it can.

Next, the effect of the storage system 1 will be described.
The first effect is that the time required for replication can be saved.
In the storage system 1, the data comparison circuit 102 (202) verifies the identity verification data (check code or file update table area data) between the master volume 120 and the replication volume 220, and if this data is the same, This is because by deleting the difference map, it is not necessary to copy data that is not originally required.
The second effect is that the line cost for relaying between the disk array devices 11 and 21 can be saved.
Due to the above characteristics, the data transmitted / received between the disk array devices 11 and 21 via the network 2 is only the identity verification data. In the replication based on the conventional method, the master / replication volumes that have flowed to the network 2 match. This is because it is no longer necessary to send the data that is being sent.

1 is a functional block diagram of a storage system that is an embodiment of the present invention. FIG. It is a figure explaining operation | movement of a storage system. It is a figure explaining operation | movement of a storage system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Storage system 2 Network 10 Primary host 11, 21 Disk array apparatus 12 Primary site 20 Secondary host 22 Secondary site 100, 200 Difference map creation circuit 101, 201 Check code creation circuit 102, 202 Data comparison circuit 110, 210 Memory 111, 211 Difference map 112, 212 Check code 120 Master volume 220 Replication volume

Claims (12)

A primary storage device including a master volume, and a secondary storage device connected to the mast storage device via a communication line and including a replication volume of the master volume, and using the difference map, the master volume and the replication volume In a storage system that performs replication
The primary storage device and the secondary storage device are respectively
When the primary storage device receives a verification instruction from the host device, the comparison target range data designated by the host device is read from the master volume or replica volume of the own device, and a check code for this data is created to create a memory. A check code creation means for storing
One of the primary storage device and the secondary storage device is
Comparing the check code read from the memory of its own device with the check code of the other device acquired via the communication line, and comprising data comparison means for notifying the comparison result to the other,
The primary storage device and the secondary storage device are respectively
A difference map update unit that updates a difference map corresponding to the comparison target range to indicate that there is no difference between the master volume and the replication volume when the check code of the own device and the other device matches;
A storage system characterized by having A primary storage device having a master volume, and a secondary storage device connected to the mast storage device via a communication line and having a replica volume of the master volume, and using the difference map, the master volume and the replica volume In a storage system that synchronizes
One of the primary storage device and the secondary storage device is
When the primary storage device receives a verification instruction from the host device, it compares the data read from the file update table storage area of its own device with the data of the file update table storage area of the other device acquired via the communication line Data comparison means for notifying the other side of the comparison result,
The primary storage device and the secondary storage device are respectively
When the data in the file update table storage area of the own device and the other device match, a difference map corresponding to the verified range designated by the host device is shown to indicate that there is no difference between the master volume and the replication volume. Data comparison means to update
A storage system characterized by having   3. The check code comparison unit, the data comparison unit, and the difference map update unit operate immediately after a pair of the master volume and the replication volume is constructed. Storage system.   4. The check code comparison unit, the data comparison unit, and the difference map update unit operate immediately after the replication volume is restored by backup of the replication volume. The storage system described in one. A difference map is used between one primary storage device having a master volume and a secondary storage device connected to the mast storage device via a communication line and having a duplicate volume of the master volume, and the master volume and the master volume In the replication method of the logical volume that synchronizes the replication volume,
When the primary storage device receives a verification instruction from the host device, the primary storage device and the secondary storage device read data in the comparison range designated by the host device from the master volume or replica volume of the own device. A check code creation process for creating a check code for this data and storing it in a memory;
A data comparison step in which one of the primary storage device and the secondary storage device compares the check code read from the memory of its own device with the check code of the other device acquired via the communication line and notifies the comparison result to the other When,
When the primary storage device and the secondary storage device match the check codes of the own device and the other device, a difference map corresponding to the comparison target range is indicated to indicate that there is no difference between the master volume and the replication volume. A difference map update process to update;
A logical volume replication method characterized by comprising: A difference map is used between one primary storage device having a master volume and a secondary storage device connected to the mast storage device via a communication line and having a duplicate volume of the master volume, and the master volume and the master volume In the replication method of the logical volume that synchronizes the replication volume,
When the primary storage device receives a verification instruction from the host device, one of the primary storage device and the secondary storage device is read from the file update table storage area of the own device and acquired via the communication line A data comparison step of comparing the data in the file update table storage area of the other device and notifying the comparison result to the other;
When the primary storage device and the secondary storage device have the same data in the file update table storage area of the own device and the other device, a difference map corresponding to the verified range designated by the host device is displayed on the master volume and the master volume. A difference map update step for updating to indicate that there is no difference in the replication volume;
A logical volume replication method characterized by comprising:   The said check code comparison process, the said data comparison process, and the said difference map update process are performed immediately after the pair of the said master volume and the said replication volume are constructed | assembled, The claim 5 or Claim 6 characterized by the above-mentioned. Logical volume replication method.   8. The check code comparison step, the data comparison step, and the difference map update step are performed immediately after the replication volume is restored by backup of the replication volume. The logical volume replication method described in 1. A difference map is used between one primary storage device having a master volume and a secondary storage device connected to the mast storage device via a communication line and having a duplicate volume of the master volume. In the logical volume replication program that synchronizes the replication volume,
When the primary storage device receives a verification instruction from the host device, the primary storage device and the secondary storage device are read from the master volume or replica volume of the device's own master volume or replication volume, and the comparison range data designated by the host device is read, Create a check code for this data and store it in memory.
Data comparison processing for comparing one of the primary storage device and the secondary storage device with a check code read from the memory of its own device and a check code of another device acquired via the communication line, and notifying the other of the comparison result The
When the check codes of the own device and the other device match the primary storage device and the secondary storage device, a difference map corresponding to the comparison target range is indicated to indicate that there is no difference between the master volume and the replication volume. Difference map update process to update,
A logical volume replication program characterized by being executed. A difference map is used between one primary storage device having a master volume and a secondary storage device connected to the mast storage device via a communication line and having a duplicate volume of the master volume, and the master volume and the master volume In the logical volume replication program that synchronizes the replication volume,
When the primary storage device receives a verification instruction from the host device, the data read from the file update table storage area of the own device is acquired via one of the primary storage device and the secondary storage device via the communication line. A data comparison process for comparing the data in the file update table storage area of the other device and notifying the comparison result to the other,
When the data in the file update table storage area of the own device and the other device matches the primary storage device and the secondary storage device, a difference map corresponding to the verified range designated by the host device is displayed on the master volume and the secondary storage device. The difference map update process that updates to indicate that there is no difference in the replication volume,
A logical volume replication program characterized by being executed.   11. The check code comparison process, the data comparison process, and the difference map update process are executed immediately after a pair of the master volume and the replication volume is constructed. Logical volume replication program.   12. The check code comparison process, the data comparison process, and the difference map update process are executed immediately after the replication volume is restored by backup of the replication volume. The logical volume replication program described in one.

Images