WO2018078718A1

WO2018078718A1 - Storage system

Info

Publication number: WO2018078718A1
Application number: PCT/JP2016/081560
Authority: WO
Inventors: 元希武井; 京介坂本; 昌宏井出; 友哉後藤
Original assignee: 株式会社日立製作所
Priority date: 2016-10-25
Filing date: 2016-10-25
Publication date: 2018-05-03

Abstract

A first volume of a first storage device and a second volume of a second storage device form an asynchronous copy pair. The second storage device includes first management information that is for managing information about host write data and that is written in the second volume according to a host overwriting write command from a host, receives a journal related to the second volume from the first storage device, and writes journal data in the second volume when the latest time associated with write data of a write destination identical to the journal data is earlier than the time associated with the journal data in the first management information.

Description

Storage system

The present invention relates to a storage system.

As a background art of the present disclosure, for example, there is JP-A-2007-86972. Japanese Patent Application Laid-Open No. 2007-86972 discloses a technology capable of realizing control including efficient duplexing recovery with reduced host load and user trouble in a storage device capable of dual writing. Specifically, the following matters are disclosed.

∙ Double-write data from the host to the primary and secondary storage device data volumes. Each storage device holds information including update data from the host as a journal (update data history information). When one of the two systems subject to dual writing becomes faulty, when returning to the duplexed state, the storage device communicates with the journal without going through the host, and A part of the data held after the most recent synchronization point necessary for volume synchronization is transferred to the recovery target system by data copy, and then the volume is restored and synchronized ( For example, summary).

Also, a remote copy system that backs up data using asynchronous remote copy technology is known. The system stores the same data in two storage devices installed at sufficiently separated primary and secondary sites. When the data in the primary storage device volume (primary volume) is updated, the update is asynchronously reflected in the secondary storage device volume (secondary volume) by remote copy. As a result, data consistency between the two storage apparatuses is ensured.

JP 2007-86972 A

When both the host dual write function and asynchronous remote copy are implemented in the storage device, the secondary storage device puts the write data transferred from the host and the copy data transferred from the primary storage device into one secondary volume. It will write against.

In other words, the secondary storage device uses data transferred by two different processes: a host I / O process that operates in synchronization with data transfer from the host, and a copy process that operates asynchronously with data transfer from the host. Update the secondary volume.

It is important for the secondary storage device to update the secondary volume in the correct order with data transferred by two different processes. In other words, it is important to avoid updating already stored data with older data in the secondary volume.

However, depending on the system configuration and operation, the data transmission order from the host may not match the data reception order in the secondary storage device. Therefore, there is a demand for a technique that can reduce the possibility of data inconsistency in a volume pair in a system that implements both the host dual write function and asynchronous remote copy.

One aspect of the present invention is a storage system including a first storage device that provides a first volume to a host, and a second storage device that provides a second volume to the host, The first volume and the second volume form an asynchronous copy pair, and the second storage device writes a host double write to the first volume and the second volume from the host. Including first management information for managing host write data information written to the second volume in response to a command. The host write data information includes the write destination of the host write data and the host write data. Receiving a journal for the second volume from the first storage device, including the time associated with The journal includes journal data, a write destination of the journal data, and a time associated with the journal data, and is associated with write data of the same write destination as the journal data in the first management information. If the latest time is older than the time associated with the journal data, the journal data is written to the second volume, and in the first management information, the same write destination host write data as the journal data If the latest time associated with is newer than the time associated with the journal data, it decides not to write the journal data to the second volume.

According to one aspect of the present invention, it is possible to reduce the possibility of data inconsistency in a volume pair in a system that implements a host double write function and an asynchronous remote copy function.

The structural example of a computer system is shown typically. The structural example of a management apparatus is shown typically. An example of the composition of a storage device is shown typically. An example of a control program executed by the storage apparatus is shown. The structural example of copy pair management information is shown. The structural example of secondary journal (SJNL) management information is shown. The structural example of comparison journal (CJNL) management information is shown. An example of a journal volume setting image displayed by the management apparatus is shown. An example of a remote copy pair setting image displayed by the management apparatus is shown. 6 shows a flowchart of processing of a write command from a host computer by the primary storage apparatus. 6 shows a flowchart of processing of a write command from the host computer by the secondary storage device. The structural example of a journal transfer command is shown. The structural example of the return value (response) with respect to a journal transfer command is shown. 5 shows a flowchart of processing of the secondary storage device in journal transfer from the primary storage device to the secondary storage device. 7 shows a flowchart of processing for reflecting journal data in a secondary volume in a secondary storage apparatus.

Hereinafter, embodiments of the present invention will be described. For clarity of explanation, the following description and drawings are omitted and simplified as appropriate. For the sake of clarity, duplicate explanation is omitted as necessary.

FIG. 1 schematically shows a configuration example of a computer system of this embodiment. The computer system includes a system at the primary site 10 and a system at the secondary site 20. The secondary site 20 is physically located away from the primary site 10. A primary storage device (also referred to as MDKC) 120 is installed at the primary site 10. A secondary storage device (also referred to as RDKC) 220 is installed at the secondary site 20.

The host computer 100 on which the business program runs, the primary storage device 120, and the secondary storage device 220 are connected by a data network 161. The platform of the host computer 100 is an open system or a mainframe.

The data network 161 is a data communication network, and is a SAN (Storage Area Network) in this configuration. The data network 161 may be a network other than SAN as long as it is a data communication network, for example, an IP network.

The host computer 100, the

storage devices

120 and 220, and the management device 140 are connected by a management network 162. The management network 162 is an IP network in this configuration. The management network 162 may be a network other than the IP network, for example, a SAN, as long as it is a network for management data communication. The data network 161 and the management network 162 may be the same network.

1, the primary storage apparatus 120 includes three primary volumes (PVOL) 121A to 121C and a primary journal volume (PJVOL) 122. The secondary storage apparatus 220 includes three secondary volumes (SVOL) 221A to 221C, a secondary journal volume (SJVOL) 222, and a comparison journal volume (CJVOL) 223.

The primary volume and the secondary volume (data volume) store user data from the host computer 100. The journal volume stores the update history of writes to the data volume.

The primary storage device 120 and the secondary storage device 220 are communicably connected via an inter-storage network 163. The primary storage device 120 and the secondary storage device 220 execute asynchronous remote copy of the data volume. Asynchronous remote copy returns an I / O completion response to the host computer 100 after writing data to the primary volume and before completing writing to the secondary volume. Note that in the following description, writing to a volume means writing to a cache or a storage drive unless otherwise specified.

The primary storage apparatus 120 stores the write data received from the host computer 100 in one of the primary volumes 121A to 121C and the primary journal volume 122. User data stored in the primary journal volume 122 is also referred to as journal data. The primary storage device 120 transfers the journal data of the primary journal volume 122 to the secondary storage device 220.

Secondary storage device 220 stores the journal data received from primary storage device 120 in secondary journal volume 222. The secondary storage device 220 stores the journal data stored in the secondary journal volume 222 in one of the secondary volumes 221A to 221C. As a result, the data of the primary volumes 121A to 121C can be copied to the secondary volumes 221A to 221C. Data copy between the storage apparatuses is executed asynchronously with the write from the host computer 100.

Remote copy is performed between data volumes that make up a copy pair. For example, the primary volume 121A constitutes a copy pair with the secondary volume 221A, the primary volume 121B constitutes a copy pair with the secondary volume 221B, and the primary volume 121C constitutes a copy pair with the secondary volume 221C.

In the example of FIG. 1, the consistency group A includes primary volumes 121A to 121C, a primary journal volume 122, secondary volumes 221A to 221C, a secondary journal volume 222, and a comparison journal volume 223. The number of various volumes included in the consistency group varies depending on the design. For example, a consistency group can include multiple primary, secondary, or comparison journal volumes.

Asynchronous remote copy matches the copy order to the secondary volume with the data write order (data update order) to the primary volume. As will be described later, a partial copy to the secondary volume may be omitted. Order matching is maintained in the consistency group. That is, the data write order to the primary volume of the consistency group in the primary storage apparatus 120 matches the write order (copy order) to the secondary volume of the consistency group in the secondary storage apparatus 220.

In the example of FIG. 1, the write data write order to the secondary volumes 221A to 221C matches the host data write order to the primary volumes 121A to 121C. The system manages the data write order by the sequential number and the time stamp.

In this embodiment, the history information of one write is called a journal. As will be described later, the journal stored in the primary and secondary journal volumes is a set of journal data (write data) and its management information. The journal stored in the comparison journal volume is composed of only management information or a combination of management information and journal data.

Journal is transferred between storage devices. The journal includes a sequential number and a write time indicating the order and time of writing to the primary volumes 121A to 121C in the consistency group.

The computer system of this embodiment implements a double writing function. The host computer 100 sends the same write command and write data to each of the primary volume and the secondary volume that constitute the copy pair. The primary storage device 120 and the secondary storage device 220 each write the received write data to the target volume. Asynchronous remote copy is not executed in double writing.

The write command includes a double write flag and indicates whether or not the write command is a double write command. The primary storage apparatus 120 refers to the double write flag in the received write command and determines whether to start the asynchronous remote copy process for the received write data.

When the secondary storage device 220 receives the double write command, it writes the write data to the secondary volume and stores the write command history information (journal) in the comparison journal volume 223. The journal of the comparison journal volume 223 includes information on the write destination and write time of write data (write command). The journal of the comparison journal volume 223 may or may not include write data. By not storing write data, resources can be used efficiently.

The secondary storage apparatus 220 selects a journal from the secondary journal volume 222 in the data reflection process from the secondary journal volume 222 to the secondary volume. The secondary storage device 220 searches the comparison journal volume 223 for a journal having a write destination common to the selected journal and indicating the latest write time.

When the write time of the journal stored in the comparison journal volume 223 is later than the write time of the journal stored in the secondary journal volume 222, storing the journal data in the secondary volume is omitted. As a result, it is possible to avoid overwriting host data by double writing with copy data older than that.

FIG. 2 schematically shows a configuration example of the management device 140. In the management device 140, a program for managing the

storage devices

120 and 220 operates. The management device 140 includes a processor 201, a memory 202, a secondary storage device 203, a management interface (I / F) 204, an output device 205, and an input device 206, which are connected to each other. The output device 205 is typically a liquid crystal display device, and the input device 206 is typically a keyboard and a mouse. The administrator can acquire necessary information in the output device 205 and input necessary information with the input device 206.

The management I / F 204 is an interface device connected to the management network 162. The processor 201 realizes a predetermined function by operating according to a program stored in the memory 202. Programs and data stored in the secondary storage device 203 are loaded into the memory 302. In FIG. 2, the memory 202 stores a storage management program 210.

The storage management program 210 manages the

storage device

120 or 220. The storage management program 210 provides the user with a user interface for setting and controlling volumes and copy pairs. Examples of images in the user interface will be described later. The storage management program 310 defines the volume configuration and copy pair configuration in the

storage apparatuses

120 and 220 according to the user.

The host computer 100 can have a general computer configuration. For example, the host computer 100 includes an interface device for connecting to the data network 161 in addition to the hardware configuration similar to that of the management apparatus 140.

FIG. 3A schematically shows a configuration example of the storage device 120/220. The storage apparatus 120/220 includes a host I / F 106, a management I / F 107, a storage controller 109 (109A, 109B), at least one cache memory 110, at least one shared memory 111, and storage drives 113 (113A to 113D). .

Suppose these hardware configurations are redundant. These elements are interconnected by a bus 112. Among these components, a set of the host I / F 106, the management I / F 107, the storage controller 109, the cache memory 110, the shared memory 111, and the bus 112 may be referred to as a storage controller.

The host I / F 106 is an interface device used for the storage apparatus 120/220 to communicate with the host computer 100. Commands (read command, write command, etc.) issued for the host computer 100 to access the volume arrive at the host I / F 106. The storage apparatus 120/220 returns information (for example, read data or write response) from the host I / F 106 to the host computer 100.

The management I / F 107 is an interface device for the storage apparatus 120/220 to communicate with the management apparatus 140. A command from the management apparatus 140 arrives at the management I / F 107. The storage device 120/220 returns information from the management I / F 107 to the management device 140.

The cache memory 110 is composed of a volatile or non-volatile memory, and temporarily stores data read / written to / from the storage drive 113. The shared memory 111 is configured by a volatile or non-volatile memory, and stores programs that operate on the storage controller 109, configuration information, and the like.

The storage controller 109 is a package board having a processor 119 and a local memory 118. The processor 119 is a device that executes a program for performing various controls of the storage apparatus 120/220. The local memory 118 is used for temporarily storing programs executed by the processor 119 and information used by the processor 119. The number of storage controllers 109 varies depending on the design.

The cache memory 110 is used to temporarily store write data for the volume (storage drive 113) or read data from the volume (storage drive 113). The shared memory 111 is a storage area for storing management information used by the storage controller 109 (the processor 119 thereof). Unlike the local memory 118, the cache memory 110 and the shared memory 111 can be accessed from the processor 119 of any storage controller 109.

The storage drive 113 is a storage device having a nonvolatile storage medium for finally storing write data from the host 101. A plurality of storage drives 113 may constitute one RAID group. The storage drive 113 can use a non-volatile semiconductor such as a magnetic disk or a flash memory as a storage medium. An example of the storage drive is an SSD (Solid State Drive). The storage drive 113 provides one or more logical storage areas (logical volumes) based on the RAID group.

FIG. 3B shows an example of a control program executed by the storage apparatus 120/220. The control program is stored in the local memory 118 and includes an input / output control program 181, a copy control program 182, a journal request program 183, a journal transfer program 184, and a secondary volume reflection program 185.

Note that one storage device can function as a primary storage device and a secondary storage device. FIG. 3B shows a program when the storage apparatus operates as both types of storage apparatuses. A storage device that functions only as one type of storage device of the primary storage device or the secondary storage device may have only the programs and information required by the one type of storage device.

The input / output control program 181 processes an I / O command (read command or write command) from the host computer 100. The input / output control program 181 generates a journal from the write command and the write data, and stores the journal in the journal volume 122.

The copy control program 182 controls the copy pair according to a command from the management apparatus 140 or the host computer 100. The journal request program 183 is executed in the secondary storage device, and requests untransferred journal data from the primary storage device. The journal request program 183 operates asynchronously with the input / output control program 181.

The journal transfer program 184 is executed in the primary storage device and transfers journal data from the secondary storage device in response to a request. The secondary volume reflection program 185 is executed in the secondary storage device. The secondary volume reflection program 185 reflects (stores) the journal data stored in the secondary journal volume to the secondary volume based on the management information stored in the secondary journal volume.

The processor operates as a functional unit (means) that realizes a predetermined function by operating according to a program. Therefore, in the present embodiment and other embodiments, the description with the program as the subject may be the description with the processor, apparatus, or system that executes the program as the subject. In this embodiment and other embodiments, the information used by the system may be expressed in any data structure without depending on the data structure.

FIG. 4 shows a configuration example of the copy pair management information 400. The copy pair management information 400 is stored in the shared memory 111 of each of the

storage apparatuses

120 and 220. The copy pair management information 400 includes information for managing each copy pair. In the example of FIG. 4, the copy pair management information 400 is represented by a table, and has a plurality of columns.

A plurality of columns are a copy pair ID column 401, a consistency group (CTG) ID column 402, a volume (VOL) ID column 403, a partner device ID column 404, a partner volume (VOL) ID column 405, a pair status column 406, a volume (VOL) type column 407 is included.

The plurality of columns further include a journal volume (JVOL) ID column 408, a partner journal volume (JVOL) ID column 409, a comparison journal (CJNL) flag column 410, and a comparison journal volume (CJVOL) ID column 411.

The copy pair ID column 401 indicates the ID of each copy pair. The consistency group ID column 402 indicates the ID of the consistency group that includes the copy pair. A volume ID column 403 indicates the volume ID of the own device in the copy pair. For example, when the copy pair management information 400 is stored in the primary storage apparatus 120, the volume ID column 403 indicates the ID of the primary volume of the primary storage apparatus 120.

The counterpart device ID column 404 indicates the ID of the storage device that stores the other volume of the copy pair. The partner volume ID column 405 indicates the ID of the other volume of the copy pair. The pair status column 406 indicates the pair status of the copy pair. An example of the pair status will be described later. The volume (VOL) type column 407 indicates the type of the volume of the own device in the copy pair, and indicates either the primary type or the secondary type.

The journal volume ID column 408 indicates the ID of the journal volume of the own device used for copy data transfer in the copy pair. The partner journal volume ID column 409 shows the ID of the journal volume of the partner device used for copy data transfer in the copy pair.

The comparison journal flag column 410 indicates whether or not double writing is applied to the copy pair. If the comparison journal flag indicates valid, a comparison journal volume for the copy pair is defined. The comparison journal volume (CJVOL) ID column 411 indicates the ID of the comparison journal volume used in duplex writing for the copy pair.

In the following, it is assumed that the pair statuses of all copy pairs included in one consistency group are the same.

Explains the pair status of asynchronous remote copy pair. The pair status of the copy pair is, for example, a DUPLEX state, a SUSPEND state, a SIMPLEX state, or a FAILURE state. In the DUPLEX state, the update of the primary volume is reflected in the secondary volume. In the SUSPEND state, the copy process is temporarily stopped. In the FAILURE state, remote copy is impossible due to a communication failure or the like. In FAILURE, remote copy is impossible due to a communication failure or the like. In the SIMPLEX state, the copy pair is released.

The copy control program 182 performs copy pair control processing. Examples of processing executed in copy control by the copy control program 182 are creation, suspension, resync, pair deletion, and status acquisition. The host computer 100 or the management apparatus 140 can request the copy control program 182 for any of the above processes. In the copy pair creation process, the copy control program 182 creates a new copy pair. The copy pair changes from the SIMPLEX state to the DUPLEX state.

The temporary stop process temporarily stops data copying from the primary volume to the secondary volume. The copy pair changes from the DUPLEX state to the SUSPEND state. When receiving the instruction for the suspension processing, the copy control program 182 does not reflect the data change of the primary volume to the secondary volume thereafter. In one example, the copy control program 182 stops the data copy between the two volumes after copying the latest update of the primary volume to the secondary volume.

The resync process indicates a process of synchronizing the primary volume data and the secondary volume data by synchronizing the primary volume and the secondary volume. The copy pair changes from the FAILURE state or the SUSPEND state to the DUPLEX state. In an example of the resync process, the copy control program 422 matches the two volume data and then restarts the copy process between them.

The pair deletion process is to cancel the pair created by the creation process. The copy pair changes from a state other than the SIMPLEX state to the SIMPLEX state. When receiving a request for status acquisition processing, the copy control program 182 transmits copy pair status information to the request source. Among the volumes included in the

storage apparatuses

120 and 220, the status information of the volumes constituting the copy pair is included in the copy pair management information 400. The copy control program 182 updates the copy pair management information 400 when executing the creation process, the resync process, and the primary stop process.

FIG. 5 shows a configuration example of secondary journal (SJNL) management information 500. The secondary journal management information 500 is stored in the secondary journal volume 222 and updated by the journal request program 183. The secondary journal management information 500 includes information for managing journal data transferred from the primary journal volume to the secondary journal volume.

In the example of FIG. 5, the secondary journal management information 500 is represented by a table and has a plurality of columns. The plurality of columns include a consistency group ID column 501, a journal volume address column 502, a write address column 503, a sequential number column 504, a write time column 505, a comparison journal flag column 506, and an entry flag column 507.

The consistency group ID column 501 indicates the consistency group ID of a copy pair in which journal data is stored. The in-journal volume address column 502 indicates the address of the secondary journal volume in which the journal data indicated by the entry is stored. A set of one entry of the secondary journal management information 500 and journal data indicated by the entry is one journal. Instead of the address where the journal data is stored, journal data may be stored.

The write address column 503 directly or indirectly indicates the storage location of journal data. The write address column 503 indicates, for example, the ID of the secondary volume and the in-volume address (for example, composed of a set of a start address and an end address). The write address column 503 indicates the ID of the primary volume and the address in the volume, and the address may be converted when writing to the secondary volume. A sequential number column 504 indicates a sequential number assigned to journal data. The sequential number indicates the write order within the consistency group.

The write time column 505 indicates the time associated with writing the journal data (write data) to the primary volume. For example, the write time column 505 indicates the time when the corresponding write command is received or the time when the write data is written to the primary volume. The primary storage apparatus 120 gives the write time to journal data (write data). The write time may be given by the host computer 100. For example, the host computer 100 includes the write command issue time in the write command.

The comparison journal flag column 506 indicates whether dual writing is applied to the secondary volume in which journal data is stored and a comparison journal volume is allocated. When the comparison journal flag indicates valid, dual writing is applied to the secondary volume in which journal data is stored. An entry flag column 507 indicates whether an entry is valid or invalid. The entry flag indicates validity when an entry is added, and indicates invalidity after the journal data is reflected on the secondary volume.

Primary journal management information (not shown) can also be represented by the same table configuration as the secondary journal management information 500. The primary journal management information is stored in the primary journal volume and updated to the input / output control program 181.

In the primary journal management information, the journal volume address column indicates an address in the primary journal volume 122. The write address column indicates, for example, the address of the secondary volume. The primary storage apparatus 120 can identify the secondary volume storing journal data with reference to the copy pair management information 400. The write address column may indicate the address of the primary volume.

FIG. 6 shows a configuration example of the comparison journal (CJNL) management information 600. The comparison journal management information 600 is stored in the comparison journal volume 223 and updated to the input / output control program 181 and the secondary volume reflection program 185. The comparison journal management information 600 includes information for managing write data stored in the secondary volume by double writing.

In the example of FIG. 6, the comparison journal management information 600 is represented by a table and has a plurality of columns. The plurality of columns include a consistency group ID column 601, a write address column 602, a write time column 603, and an entry flag column 604. One entry indicates data written by one double writing.

The consistency group ID column 601 indicates the ID of the consistency group that includes the write data storage destination secondary volume. The write address column 602 indicates a write destination of the write data, and indicates, for example, an ID of the secondary volume and an address in the volume (for example, a set of a start address and an end address).

The write time column 603 indicates the time associated with writing the write data to the secondary volume. The write time for dual writing is determined by the same method as the write time given to the asynchronous remote copy journal data. The value of the write time column 603 is determined by the same method as the value of the write time column 505.

For example, the write time column 603 indicates the time when the corresponding write command is received or the time when the write data is written to the secondary volume. The secondary storage device 120 gives the write time to the write data that is written twice. The write time may be given by the host computer 100. For example, the host computer 100 includes the write command issue time in the write command.

The entry flag column 604 indicates whether the entry is valid or invalid. The entry flag indicates validity when an entry is added, and the entry flag of an unnecessary entry indicates invalid after the reflection processing to the secondary volume. When the comparison journal volume 223 also stores write data, the comparison journal management information 600 indicates a write data storage destination address in the comparison journal volume 223 or includes write data.

FIG. 7 shows an example of a journal volume setting image displayed by the management apparatus 140. The storage management program 210 displays a journal volume setting image on the output device 205. The user inputs necessary information via the input device 206.

The storage management program 210 accepts designation of journal volumes and their types included in the consistency group in the journal volume setting image. The “standard” journal volume is a primary or secondary journal volume, and the “host double write” journal volume is a comparison journal volume. The comparison journal volume is set only in the secondary storage device. The input information is transferred to the primary storage device 120 and the secondary storage device 220 by the storage management program 210 and reflected in the management information including the copy pair management information 400.

FIG. 8 shows an example of a remote copy pair setting image displayed by the management apparatus 140. The storage management program 210 displays the remote copy pair setting image on the output device 205.

The storage management program 210 accepts designation of the remote copy type and the consistency group including the remote copy pair in the remote copy pair setting image. The storage management program 210 further accepts designation of a primary volume and a secondary volume that constitute a remote copy pair, and a journal volume that stores journal data or dual-write management information. The comparison journal volume is not specified for copy pairs to which dual writing is not applied.

The processing example of this computer system will be described below. FIG. 9 shows a flowchart of processing of a write command from the host computer 100 by the primary storage apparatus 120. The primary storage apparatus 120 receives a write command and write data from the host computer 100 (S101). The write command and write data are temporarily stored in the buffer of the host I / F 106.

The input / output control program 181 acquires and analyzes the write command (S102). The input / output control program 181 identifies the target volume of the write command, and refers to the copy pair management information 400 to determine whether the target volume is a primary volume of an asynchronous remote copy pair (S103).

When the target volume of the write command is not the primary volume of the asynchronous remote copy pair (S103: NO), the I / O control program 181 writes the write data to the target volume (S106).

When the target volume of the write command is the primary volume of the asynchronous remote copy pair (S103: YES), the I / O control program 181 determines whether the write command is double writing (S104). When the double write flag of the write command indicates valid (S104: YES), the input / output control program 181 sets the host double write flag to ON on the local memory 118 (S105), and writes the write data to the target volume. Writing is performed (S106). If the write command is not double writing (S104: NO), step S105 is skipped.

After writing the write data to the target volume (S106), the I / O control program 181 returns a write completion response to the host computer 100 (S107). The input / output control program 181 refers to the host double write flag in the local memory 118 (S108). When the host double write flag is ON (S108: YES), the input / output control program 181 ends this flow. Thereby, it is possible to omit generation of an unnecessary journal in the case of double writing.

When the host dual write flag is OFF (S108: NO), the I / O control program 181 refers to the copy pair management information 400 and indicates whether there is a pair volume of the write target volume and the copy pair including the write target volume. The state is identified (S109).

When the pair volume exists and the pair status is the DUPLEX status (S109: YES), the I / O control program 181 creates a new journal corresponding to the received write command and stores it in the primary journal volume 122 (S110). ). Data copying is not required when there is no pair volume or in another pair status. A journal (journal data and management information) can be stored in the primary journal volume 122 when writing to the primary volume of the asynchronous remote copy pair alone.

As described above, the I / O control program 181 refers to the copy pair management information 400 when creating a new journal, and checks the consistency group of the target primary volume, the secondary volume of the pair, and whether or not dual writing is applied. it can. The input / output control program 181 determines the write time based on the reception of the write command, storage in the primary volume, or information included in the write command. The entry flag of the added entry indicates validity.

If the pair status is not DUPLEX (S109: NO), the input / output control program 181 ends this flow. According to this flow, asynchronous remote copy processing can be executed appropriately and efficiently in accordance with the pair status of the remote copy pair.

FIG. 10 shows a flowchart of processing of a write command from the host computer 100 by the secondary storage apparatus 220. The secondary storage apparatus 220 receives a write command and write data from the host computer 100 (S151). The write command and write data are temporarily stored in the buffer of the host I / F 106.

The input / output control program 181 acquires and analyzes the write command (S152). The input / output control program 181 identifies the target volume of the write command, refers to the copy pair management information 400, and determines whether the target volume is a secondary volume of an asynchronous remote copy pair (S153).

When the target volume of the write command is not the secondary volume of the asynchronous remote copy pair (S153: NO), the I / O control program 181 writes the write data to the target volume (S154) and sends a write completion response to the host computer 100. Return (S155), and the flow ends.

When the target volume of the write command is a secondary volume of an asynchronous remote copy pair (S153: YES), the I / O control program 181 determines whether the write command is double writing (S156). If the write command is not double writing (S156: NO), the input / output control program 181 proceeds to step S154.

When the double write flag of the write command indicates valid (S156: YES), the input / output control program 181 sets the host double write flag to ON on the local memory 118 (S157). Next, the input / output control program 181 refers to the copy pair management information 400 to identify the status of the copy pair including the write target volume (S158, S159).

If the pair status is the SUSPEND status or the SIMPLEX status (S158: YES), the input / output control program 181 proceeds to step S154. If the pair status is neither the SUSPEND status, the SIMPLEX status, or the DUPLEX status (S158: NO, S159: NO), that is, if the pair status is the FAILURE status, the I / O control program 181 reports an error to the host computer 100. (S160).

When the pair status is the DUPLEX status (S158: NO, S159: YES), the input / output control program 181 refers to the host double write flag in the local memory 118 (S161). When the host double write flag is OFF (S161: NO), the input / output control program 181 reports an error to the host computer 100 (S160).

As a result, in the DUPLEX state (state in which asynchronous copying is performed from the primary volume to the secondary volume), it is possible to avoid data inconsistency in the copy pair due to a write command (write command that is not host-host duplex writing) for only the secondary volume. .

If the host double write flag is ON (S161: YES), the I / O control program 181 writes the write data to the secondary volume (S162), and returns a write completion response to the host computer 100 (S163). The input / output control program 181 further adds a new entry corresponding to the write command (write data) to the comparison journal management information 600. The entry flag of the added entry indicates validity.

The I / O control program 181 can identify the consistency group of the write target volume with reference to the copy pair management information 400. The write address is indicated in the write command. The input / output control program 181 determines the write time based on the reception of the write command, storage in the primary volume, or information included in the write command. The entry flag of the added entry indicates validity.

According to the above flow, other types of write commands that can appropriately update the comparison journal management information 600 in accordance with the double write command for the secondary volume can be appropriately processed.

Next, journal transfer between storage devices will be described. In response to the journal transfer command from the secondary storage apparatus 220, the primary storage apparatus 120 transmits a return value.

FIG. 11 shows a configuration example of the journal transfer command 250. The journal transfer command 250 is a command for transferring the journal of the primary storage device 120 to the secondary storage device 220. The journal request program 183 of the secondary storage device 220 creates a journal transfer command 250 and issues it to the primary storage device 120.

The journal transfer command 250 includes an instruction code field 251, a consistency group ID field 252, a sequential number field 253 of a journal already reflected in the secondary volume, and a received sequential number field 254.

The instruction code field 251 stores the instruction type of the command, that is, data indicating a journal transfer command. The consistency group ID field 252 stores the consistency group ID of the journal requested to be transferred. The field 254 stores the last written journal, that is, the sequential number of the latest reflected journal in the consistency group to which the transfer requested journal belongs.

The received sequential number section field 255 is the sequential number of the journal last received by the secondary storage system 220 in the consistency group to which the journal requested to be transferred belongs. That is, it is the sequential number of the latest journal held in the secondary journal management information 500.

The journal request program 183 refers to the secondary journal management information 500 and generates a journal transfer command 250. The journal request program 183 inputs the latest sequential number indicated by the secondary journal management information 500 in the received sequential number field 254. The journal request program 183 inputs the latest sequential number indicating that the entry flag is invalid in the sequential number field 253 of the journal already reflected in the SVOL.

The journal transfer program 424 of the primary storage device 120 transmits a return value including a journal in response to the journal transfer command 250 from the secondary storage device 220.

FIG. 12 shows a configuration example of a return value (response) 350 for the journal transfer command. The journal transfer program 424 of the primary storage device 120 transmits a return value 350 to the secondary storage device 220 as a response to the journal transfer command.

The return value 350 includes a consistency group ID field 351, a journal field 352, and a latest sequential number field 353. The consistency group ID field 351 stores the consistency group ID of the requested journal. This value is the same as the value in the consistency group ID field 252 in the journal transfer command 250.

The journal field 352 includes one or more journals transferred in response to the journal transfer request command. In the example of FIG. 12, three journals are included. The number of journals stored depends on the number or data volume of untransferred journals and the design. Each journal includes journal data, a write address, a sequential number, a write time, and a comparison journal flag.

The journal transfer program 184 acquires a prescribed number of journals from the primary journal volume 122 from the sequential number next to the number indicated by the received sequential number field 254.

The journal transfer program 184 inputs journal data acquired from the primary journal volume 122 and information acquired from primary journal management information (not shown) in the journal field 352. As described above, the primary journal management information has the same configuration as the secondary journal management information 500, and stores information on the write address, sequential number, write time, and comparison journal flag.

The journal transfer program 184 inputs the latest sequential number when the journal transfer command 250 is received in the primary journal management information in the latest sequential number field 353.

The journal transfer program 184 changes the entry flag of the journal data entry already reflected in the secondary volume to invalid. The journal transfer program 184 changes the entry flag to invalid in the entry of the sequential number before the sequential number indicated by the sequential number field 253 of the journal that has been reflected in the secondary volume of the journal transfer command 250.

The journal transfer program 184 may change the entry flag of the transferred journal data entry to invalid. The journal transfer program 184 changes the entry flag to invalid in the entry of the sequential number before the latest sequential number in the return value 350 transferred this time. The journal data of the invalidated entry is deleted.

The journal request program 183 of the secondary storage apparatus 220 extracts a journal from the received return value 350 and stores it in the secondary journal volume 222. The journal request program 183 converts the journal if necessary. For example, when the write address indicates the primary volume, the journal request program 183 converts the write address and stores it in the secondary journal management information 500.

FIG. 13 shows a flowchart of processing of the secondary storage apparatus 220 in journal transfer from the primary storage apparatus 120 to the secondary storage apparatus 220. In this embodiment, journal transfer is executed asynchronously with the write from the host computer 100. In the present embodiment, journal transfer is executed for the selected consistency group.

The journal request program 183 of the secondary storage device 220 requests the primary storage device 120 to transfer an untransferred journal at a predetermined timing (S201).

When it is time to request journal transfer (S201: YES), the journal request program 183 refers to the copy pair management information 400 and confirms the pair status of the consistency group (S202). If the pair status is not the DUPLEX status (S202: NO), the data copy is unnecessary, and the journal request program 183 ends this flow.

When the pair status is the DUPLEX status (S202: YES), the journal request program 183 requests journal transfer to the primary storage device 120 (S203). The journal request program 183 generates a journal transfer command 250 and transmits it to the primary storage device 120. Details of the journal transfer command have been described with reference to FIG.

When the primary storage device 120 receives the journal transfer command, the journal transfer program 184 transmits a return value 350 of the journal transfer command to the secondary storage device 220. The secondary storage device 220 receives the return value 350 (S204). Details of the return value 350 of the journal transfer command have been described with reference to FIG.

The journal request program 183 of the secondary storage apparatus 220 stores the journal included in the received return value in the secondary journal volume 222. Specifically, the journal request program 183 stores the received journal data in the secondary journal volume 222, and adds new journal data management information to the secondary journal management information 500. The journal request program 183 may repeatedly transmit the journal transfer command 250 until the journal having the latest sequence number indicated by the return value 350 is received.

Through the above processing, the secondary storage apparatus 220 can receive the untransferred journal of the selected consistency group from the primary storage apparatus 120.

FIG. 14 shows a flowchart of the process of reflecting journal data to the secondary volume in the secondary storage apparatus 220. The secondary volume reflection program 185 repeatedly executes the process at a predetermined timing. This process is executed synchronously or asynchronously with the journal reincarnation process.

The secondary volume reflection program 185 refers to the secondary journal management information 500 and selects the oldest entry (comparison source secondary journal entry) indicating that the entry flag is valid (S301). The secondary volume reflection program 185 determines whether the comparison journal flag of the selected comparison source secondary journal entry indicates validity (S302).

When the comparison journal flag indicates invalid (S302: NO), the secondary volume reflection program 185 stores the journal data of the comparison source secondary journal entry in the secondary volume area indicated by the write address (S305). Since double writing is not applied to the copy pair, comparison with the comparison journal management information 600 is unnecessary, and the processing is made efficient.

If the comparison journal flag is valid (S302: NO), double writing is applied to the copy pair. There is a possibility that write data that is newer than the comparison source secondary journal entry has been written to the secondary volume by host double writing.

Therefore, the secondary volume reflection program 185 searches the comparison journal management information 600 for the latest valid entry having the same address as the write address indicated by the comparison source secondary journal entry (S303). Here, it is assumed that the access by the write command or the address size in the journal is constant.

If no valid entry with the same write address exists in the comparison journal management information 600 (S304: NO), the journal data of the comparison source secondary journal entry is the latest data. The secondary volume reflection program 185 stores the journal data in the secondary volume area indicated by the write address (S305). The secondary volume reflection program 185 changes the entry flag of the comparison source secondary journal entry to invalid (S306). The journal data of the entry whose entry flag is invalidated is deleted.

When a valid entry with the same write address exists in the comparison journal management information 600 (S304: YES), the secondary volume reflection program 185 determines the latest write time in the comparison journal management information 600 and the comparison source for each overlapping address area. The write time of the secondary journal entry is compared (S309).

When the write time of the comparison source secondary journal entry is new (S310: YES), the secondary volume reflection program 185 stores the journal data in the secondary volume area indicated by the write address (S305). The secondary volume reflection program 185 changes the entry flag of the comparison source secondary journal entry to invalid (S306).

If the write time of the comparison source secondary journal entry is old (S310: NO), the secondary volume reflection program 185 stores the entry data of the comparison source secondary journal entry without storing the journal data in the secondary volume area indicated by the write address. Is changed to invalid (S306).

When the selected comparison source secondary journal entry is not the last valid entry in the secondary journal management information 500 (S307: NO), the secondary volume reflection program 185 returns to step S302.

When the selected comparison source secondary journal entry is the last valid entry in the secondary journal management information 500 (S307: YES), the secondary volume reflection program 185 in the comparison journal management information 600 determines the last in the secondary journal management information 500. The entry flag of the entry before the entry write time is set to invalid (S311). This makes it possible to invalidate entries that are unnecessary in future comparison processing, and to make the comparison processing more efficient. In the secondary journal management information 500, the transfer time from the primary storage device 120 may be managed instead of the write time.

As described above, when data newer than the journal data transferred from the primary storage device 120 has been written to the secondary volume by the host double writing, the journal data is not stored in the secondary volume. As a result, it is possible to prevent old journal data from overwriting new host data.

As described above, since the secondary journal management information 500 includes the comparison journal flag, the secondary storage apparatus 220 can efficiently know whether double writing is applied to journal data. The above example further includes a comparative journal flag in the journal. As a result, it is possible to know whether or not the double writing is applied to the journal data more efficiently.

Instead, the comparison journal flag is omitted in the journal, and the secondary storage apparatus 220 adds the comparison journal flag to the secondary journal management information 500 with reference to the comparison journal flag column 410 of the copy pair management information 400. Also good. The comparison journal flag may be omitted in the secondary journal management information 500, and the secondary storage apparatus 220 may refer to the comparison journal flag column 410 of the copy pair management information 400.

The above example invalidates the entry by setting the entry flag of the unnecessary entry in the management information invalid, but the entry may be invalidated by deleting the unnecessary entry. The write command and journal write address range need not be constant. The secondary volume reflection program 185 searches the comparison journal management information 500 for the latest valid entry that overlaps (at least partially matches) the write address of the comparison source secondary journal entry. The comparison journal volume may not be defined. For example, the comparison journal management information 500 may be stored in the shared memory 111.

The secondary volume reflection program 185 compares the write time between the secondary journal and the comparison journal for each overlapping address range. In the secondary journal, data in an address range that does not match the address of any valid comparison journal is reflected in the secondary volume.

The above configuration example compares the time of the journal data and the host double write data in reflecting the secondary journal data, and writes the host data to the secondary volume without comparing with the journal data in the host double write. This effectively avoids overwriting new data with old data.

The secondary storage device 220 may compare the write time between the host data and the journal data in the host double writing. Normally, data transfer by asynchronous remote copy is slower than data transfer by host write. Therefore, the write times may be compared in storing journal data in the secondary volume. By comparing the write times of host data and journal data in host double writing, the reliability of data consistency can be improved.

The time included in the write command from the host computer 100 is used as the write time. The input / output control program 181 searches the secondary journal management information 500 for the latest invalid entry having the same address as the host write data address.

If newer journal data than the host write data has already been written to the secondary volume, the I / O control program 181 returns a write completion application to the host computer 100 without writing the host write data to the secondary volume. The entry of the host write data that has not been written to the secondary volume is set to invalid in the comparison journal management information 600.

The update of the secondary journal management information 500 and the processing of the entry of the host data written to the primary journal volume in the comparison journal management information 600 may be the same as the other configuration examples described above. The input / output control program 181 may search for the latest entry at the same address in the valid and invalid entries. The comparison when the addresses of the host write data and the journal data partially match is the same as in the above other configuration examples.

In addition, this invention is not limited to the above-mentioned Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

In addition, each of the above-described configurations, functions, processing units, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files for realizing each function can be stored in a memory, a recording device such as a hard disk drive or SSD, or a recording medium such as an IC card or SD card. In addition, the control lines and information lines are those that are considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. In practice, it may be considered that almost all the components are connected to each other.

Claims

A first storage device that provides a first volume to a host;
A second storage apparatus that provides a second volume to the host, the storage system comprising:
The first volume and the second volume form an asynchronous copy pair;
The second storage device
First management information for managing host write data information written to the second volume in response to a host double write command from the host to the first volume and the second volume. The host write data information includes a write destination of the host write data and a time associated with the host write data,
Receiving a journal for the second volume from the first storage device, the journal including journal data, a write destination of the journal data, and a time associated with the journal data;
In the first management information, when the latest time associated with the write data of the same write destination as the journal data is older than the time associated with the journal data, the journal data is stored in the second volume. Write to
In the first management information, when the latest time associated with the same host write data as the journal data is newer than the time associated with the journal data, the journal data is stored in the second management information. A storage system that decides not to write to a volume.
The storage system according to claim 1,
The second storage device includes second management information for managing journal information transmitted from the first storage device,
The second management information includes a write destination of the journal data and a time associated with the journal data,
Receiving a first host double write command and first host write data from the host to the first volume and the second volume;
Adding the information of the first host write data to the first management information;
A storage system that writes the first host write data to the second volume without referring to the second management information.
The storage system according to claim 1,
The second storage device
A storage that invalidates information of host write data before the time associated with the journal data in the first management information after the comparison between the first management information and the time associated with the journal data. system.
The storage system according to claim 1,
The second storage device
Indicates whether the host dual write is applied to the journal storage area for storing the journal data transmitted from the first storage device and the write destination of the journal data transmitted from the first storage device. Third management information, and
When the third management information indicates that host dual writing is applied to the write destination of the first journal data in the journal storage area, the time associated with the first journal data and the first Compared with the management information of
When the third management information indicates that host dual writing is not applied to the write destination of the first journal data, the first journal data is transferred to the target volume without referring to the first management information. To write to the storage system.
The storage system according to claim 1,
Including fourth management information for managing a status of an asynchronous copy pair of the first volume and the second volume;
The second storage device
Analyzing the first write command received from the host to determine whether the write destination of the first write command and the first write command are host double writing;
The write destination of the first write command is the second volume, the first write command is not host double writing, the fourth management information is the status of the asynchronous copy pair A storage system that returns an error to the host when indicating that it is in a state of performing asynchronous copy from one volume to a second volume.
The storage system according to claim 1,
The first storage device
Receiving a second write command and second host write data from the host to the first volume;
Analyzing the second write command;
If the second write command is a host double write command, the second host write data is transferred to the first volume without transferring the second host write data to the second volume. Write and
A storage system for transferring the second host write data to the second volume when the second write command is not a host dual write command.
A first storage device that provides a first volume to the host, and a second storage device that provides a second volume to the host, the first volume and the second volume, Is a storage system control method for forming an asynchronous copy pair,
Information on host write data written to the second volume in response to a host double write command from the host to the first volume and the second volume is sent from the second storage device. Including first management information to be managed, the host write data information includes a write destination of the host write data and a time associated with the host write data,
The second storage device receives a journal for the second volume from the first storage device, and the journal is associated with journal data, a write destination of the journal data, and the journal data. Including
In the first management information, when the latest time associated with the write data of the same write destination as the journal data is older than the time associated with the journal data, the second storage device Write journal data to the second volume,
When the latest time associated with the same write destination host write data as the journal data is newer than the time associated with the journal data in the first management information, A method of determining not to write the journal data to the second volume.
The method of claim 7, comprising:
The second storage device includes second management information for managing journal information transmitted from the first storage device,
The second management information includes a write destination of the journal data and a time associated with the journal data,
The method
The second storage device receives the first host double write command and the first host write data from the host to the first volume and the second volume;
The second storage device adds the information of the first host write data to the first management information;
A method in which the second storage device writes the first host write data to the second volume without referring to the second management information.
The method of claim 7, comprising:
After the second storage device compares the first management information and the time associated with the journal data, host write data before the time associated with the journal data in the first management information How to invalidate information.
The method of claim 7, comprising:
The second storage device has a host duplex for a journal storage area for storing journal data transmitted from the first storage device and a write destination for journal data transmitted from the first storage device. Including third management information indicating whether the writing is applied;
When the second storage device indicates that the host management write is applied to the write destination of the first journal data in the journal storage area, the third management information is added to the first journal data. Comparing the associated time with the first management information;
When the second storage device indicates that the host management write is not applied to the write destination of the first journal data, the third management information does not refer to the first management information. A method of writing first journal data to a target volume.
The method of claim 7, comprising:
The second storage device includes fourth management information for managing a status of an asynchronous copy pair of the first volume and the second volume,
The method
The second storage device analyzes the first write command received from the host and determines whether the write destination of the first write command and the first write command are host double writing;
In the second storage device, the write destination of the first write command is the second volume, the first write command is not host double write, and the fourth management information is the asynchronous A method of returning an error to the host when the status of the copy pair indicates that asynchronous copy is performed from the first volume to the second volume.
The method of claim 7, comprising:
The first storage device receives a second write command and second host write data from the host to the first volume;
The first storage device analyzes the second write command;
When the second write command is a host double write command, the first storage device transfers the second host write data without transferring the second host write data to the second volume. Write data to the first volume;
A method in which the first storage device transfers the second host write data to the second volume when the second write command is not a host double write command.