JP2000163277A - Method for data transfer between dasd controllers - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten the time needed for a copying processing by transferring write instructions to a recovery system and performing the copying processing to DASD of the recovery system after a writing process to the DASD was continuously performed. SOLUTION: Primary PRA22A issues a write instruction WD1 to secondary PRA22B through an interface and the secondary PRA22B when completing the writing processing to a secondary device as to the write instruction WD1, returns a write completion note to the primary PRA22A. Similarly, write instructions WD2 and WD3 and a writing end note are transferred between the primary PRA22A and secondary PRA22B. The primary PRA22A writes and updates the completion of all copies to the secondary device in a progress table and CA21A once deciding the completion issues a connection request to a channel 10A.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a DASD controller for controlling a DASD (Direct Access Storage Device) of a normal operation system in order to duplicate the data held in the DASD of the recovery system in order to duplicate the data in the DASD of the recovery system. And a data transfer method for transferring update data (write data) from a normal operation system to a recovery system.

[0002]

2. Description of the Related Art Even if a normal operation system for handling user data falls into an unusable state due to a disaster or the like, operation can be immediately continued using the user data without losing the user data. Therefore, generally, the same data as the user data in the normal operation system is stored in the recovery system to duplicate the user data.

[0003] Conventionally, a DASD control device for controlling a DASD (for example, a magnetic disk device) in each of two systems is connected to each other, and update data in the normal operation system is transferred from the normal operation system to the recovery system. Synchronous copy function (remote transfer function) to ensure data equivalence in DASD
Is provided. This copy function is a synchronous copy function that maintains the identity of all data to separate systems separated from each other, and provides a function in which data is not lost even if some problem occurs in the normal operation system. The above two systems are, for example, MVS (Multip
OS for large general-purpose computers (Operati
ng System), the DASD controllers
For example, it is connected by ESCON interface.

FIG. 3 is a block diagram showing a general configuration of a system having a copy function (remote transfer function) as described above. In FIG. 3, an application site indicates a normal operation system, The recovery site refers to a recovery system that holds data equivalent to the data of the primary device (DASD) of the application site. In FIG. 3, the same numbers are assigned to components that perform the same function at the application site and the recovery site, and “A” is assigned to the symbols that indicate the components belonging to the application site. “B” is attached to the reference numeral indicating the component belonging to the site.

[0005] The application site includes a host 1A, a primary control unit (DASD controller) 2A, and a primary device (DASD) 3A which are used for normal operation. The primary control unit 2A controls access to the primary device 3A in accordance with an instruction from the host 1A. The primary control unit 2A includes a CA (Channel Adapter) 21A that controls data transfer with the channel 10A of the host 1A and an interface 4 such as an ESCON interface. Via primary /
PRA (Peer-to-peer) that controls data transfer between secondary
Remote Adapter) 22A.

Similarly, the recovery site includes a host 1B used for holding equivalent data, a secondary control unit (DASD control device) 2B, and a secondary device (DASD) 3B. The secondary control unit 2B is a secondary device 3
B for controlling access to CA 21B
And a PRA 22B that controls data transfer between the primary and secondary via the interface 4.

Next, in the system configured as described above, the operation when the host 1A of the application site issues a write command to the data of the primary device 3A to the primary control unit 2A (copy function / remote transfer function). Will be described with reference to FIG. 4 and FIG. FIG. 4 shows a write command (CCW: Channel Command) issued by the host 1A.
FIG. 5 is an operation flowchart for explaining a series of data transmission / reception (conventional copy processing) between an application site and a recovery site.

A write command from the host 1 is, for example, as shown in FIG.
Is notified by a channel command sequence called CCW as shown in FIG. That is, the type of the following read / write operation is specified by the DX (Define Extent) command and the LR (Locate Record) command. Here, the write process (WD1 to WD) of the data portion for three designated records is performed.
Assume 3).

As shown in FIG. 5, the CA 21A receives the DX command and the LR command from the channel 10A of the host 1A (step A11), and then receives the first write command WD1 (step A12). (Step A13). Then, the CA 21A writes and updates the fact that a copy event to the secondary device 3B has occurred (copy factor) in a progress table (not shown; present in the main memory),
When reading the copy factor with reference to the progress table (step A14), the primary PRA 22A issues a write command WD1 to the secondary PRA 22B via the interface 4 (step A15).

When completing the write processing to the secondary device 3B at the recovery site, the secondary PRA 22B returns a write completion notification to the primary PRA 22A (step A16). Primary PRA22A
Writes and updates the completion of copying to the secondary device 3B in the progress table, and the CA 21A refers to the progress table to determine the end of the processing of the write command WD1 (Steps A17 and A18), and the channel of the host 1A. The end is notified to 10A (step A19).

Hereinafter, the processing of steps A21 to A28 is performed for the write instruction WD2, and the processing of steps A31 to A38 is performed for the write instruction WD3. here,
Steps A21-A28 correspond to steps A12-A described above.
Step A3
1 to A38 correspond to the processing of steps A12 to A19, respectively. That is, after receiving the DX command and the LR command from the channel 10A,
The data part for three designated records is the secondary device 3
The same process is repeated three times for copying to B.

The above is a series of flow of the copy process from the application site to the recovery site for one set of CCW chains. By performing the copy process as described above, the host 1
Even if A makes a change such as writing to the primary device 3A, equivalent data is always held in the primary device 3A in the application site and the secondary device 3B in the recovery site.

[0013]

The copy function between the control units (DASD control devices) 2A and 2B operates in conjunction with normal write processing. When a write process for the data of the primary device 3A occurs in the application site, the equivalent data is copied to the secondary side (recovery site side) by the internal processing of the primary control unit 2A. At this time, the host 1A of the application site It operates regardless of the presence or absence of the copy function. In the write process of the copy function, only the time required for copying to the secondary
The processing time is longer than the normal processing, and a performance difference from the normal processing is required.

That is, in the prior art, the copy processing time to the secondary takes extra time than the general normal operation processing, so that the processing time increases, and the processing time increases. The occupation time of the I / O processing also increases, which is a factor of deteriorating the processing capacity of other I / O processing. The present invention has been made in view of such a problem, and aims to reduce the time required for a recovery system to perform a copy process to a DASD, and to release a channel during the copy process to release another I / O. It is an object of the present invention to provide a data transfer method in which the performance of a copy function between DASD controllers is improved in a comprehensive manner so that the data transfer method can be used for processing.

[0015]

In order to achieve the above object, a method for transferring data between DASD controllers according to the present invention (claim 1) uses a DASD of a recovery system for transferring data held in a DASD of a normal operation system. In order to make the system redundant, the DASD controllers controlling the DASDs in these systems are connected to each other, and update data for the DASD of the normal operation system is transferred from the normal operation system to the recovery system, and A method for copying to a DASD, wherein in the normal operation system, when a plurality of write instructions that cause a copy to the DASD of the recovery system are successively issued from a higher-level device to the DASD controller,
DA of the normal operation system in response to the plurality of write instructions
After continuously performing write processing on SD, the plurality of write instructions are transferred to the recovery system, and copy processing on DASD of the recovery system is performed.

At this time, the plurality of write instructions may be combined into one write instruction and transferred to the recovery system (claim 2). In the operation system, the channel used to transmit the plurality of write commands from the higher-level device to the DASD controller may be released (claim 3).

In the above-described data transfer method between DASD controllers according to the present invention (claim 1), when a plurality of write instructions are continuously issued, a write operation corresponding to the plurality of write instructions is performed in a normal operation system. Since the transfer (copy processing) to the recovery system is performed after the processing is performed, the overlapping processing can be eliminated as compared with the related art in which the copy processing is performed every time the write processing is performed for one write instruction.

At this time, by transferring a plurality of write instructions collectively into one write instruction (claim 2), the time required for the copy processing can be shortened, and the channel occupied conventionally during the copy processing is released. By doing so (claim 3), the channel can execute other I / O processing.

[0019]

Embodiments of the present invention will be described below with reference to the drawings. DA as an embodiment of the present invention
Since the system to which the data transfer method between the SD control devices is applied is configured in the same manner as that shown in FIG. 3, a detailed description thereof will be omitted. The data transfer method according to the present embodiment can be easily realized by slightly modifying the firmware (program) for the conventional data transfer described above with reference to FIG.

In one embodiment of the present invention,
A data transfer procedure (an example of data transmission / reception between the primary control unit 2A and the secondary control unit 2B) for ensuring data equivalence in the application site and the recovery site will be described with reference to FIG. FIG. 1 is an operation flowchart for explaining a series of flows of data transmission / reception (copy processing) between the primary control unit 2A and the secondary control unit 2B upon activation from the host 1A on the application site side.

Also in this embodiment, the host 1A of the application site is connected to the primary device 3A.
Write instruction for the data in (CCW column shown in FIG. 4)
(Copy function / remote transfer function) when is issued to the primary control unit 2A. As shown in FIG. 1, the CA 21A is connected to the host 1A.
When the DX command and the LR command are received from the channel 10A (step S11), the type and number (record number) of the following read / write processing can be recognized by the command parameters of the LR command. For example, in the assumption of the CCW chain shown in FIG. 4, the content of the LR command allows the CA 21A to know in advance that three write data will be successively chained below. As a result, the CA 21A can recognize that the copy processing to the secondary is also for three records, and performs this copy processing for three records continuously after the write processing to the primary.

That is, the CA 21A subsequently receives the first write command WD1 (step S12), performs data write processing to the primary device 3A (step S13), and sends the channel 10A a write command WD1 on the primary side. (Step S14). The next write command WD2 is sent from the host 1A (channel 10A) that has received the write completion notification of the write command WD1 (step S21), and the CA 21A that receives the write command WD2 sends the data to the primary device 3A. Write processing is performed (step S2
2) Notify the channel 10A of the end of the write processing by the write instruction WD2 on the primary side (step S23). Then, the last write command WD3 is sent from the host 1A (channel 10A) that has received the write completion notification of the data WD2 (step S31).

C receiving the third write instruction WD3
After performing the data write process on the primary device 3A (step S32), the A21A recognizes that it is necessary to perform the copy process on the secondary device 3B on the recovery site side, and during the copy process, the channel 1
A disconnection request is issued to the channel 10A to disconnect from the host 1A (step S33), and a disconnection notification is received from the host 1A (channel 10A) receiving the disconnection request (step S34), thereby disconnecting from the host 1A. You.

Subsequently, the CA 21A writes and updates in the progress table that copy processing for three records needs to be performed on the secondary device 3B (copy factor), and the primary PRA 22A refers to the progress table to update the secondary device 3B. When the necessity of performing the copy process to the server is recognized (step S41), the copy process is started. That is, the primary PRA 22A first issues the write command WD1 to the secondary PRA 22B via the interface 4 (step S42). When the secondary PRA 22B completes the write processing for the write command WD1 to the secondary device 3B, the primary PRA 22A A write end notification is returned to the PRA 22A (step S43). Similarly, the primary PRA 22A and the secondary PRA 2
2B, write commands WD2 and WD3 and a write end notification are exchanged (steps S44 to S4).
7).

Then, the primary PRA 22A, which has received the write end notification for the third write instruction WD3,
The progress table is written and updated to the effect that all copy processing to the secondary device 3B has been completed, and the CA 21A determines that the copy processing for the write commands WD1 to WD3 has been completed with reference to the progress table (steps S48 and S4).
9), issue a connection request to channel 10A (step S)
50), by receiving a connection notification from the host 1A side (channel 10A) receiving the connection request (step S5).
1) Connected to host 1A. In this state, CA21
A notifies the end to the channel 10A (step S52).

As described above, this is a series of flow of the copy process from the application site to the recovery site for one set of CCW chains using the data transfer method of the present embodiment. By doing so, even if a change such as writing to the primary device 3A is made by the host 1A in the application site, equivalent data is always stored in the primary device 3A in the application site and the secondary device 3B in the recovery site. Will be retained.

As described above, according to the data transfer method of the present embodiment, the copy processing to the secondary is performed continuously (see steps S41 to S48), and
The update / reference processing of the progress table, which has been performed independently for each write instruction, can be integrated, thereby eliminating the duplication processing and increasing the processing speed. During the copy process, the channels 10A and CA22A are used.
Can be disconnected and the channel 10A can be released, so that this channel 10A can receive other I / O processing, and the performance as an apparatus can be improved.

As described above, according to the present invention, the performance of the copy function between the control units 2A and 2B can be greatly improved. It should be noted that the present invention is not limited to the above-described embodiment, and can be implemented with various modifications without departing from the spirit of the present invention. For example, in the above-described embodiment, the processing performed for each write instruction as shown in steps S42 to S47 of FIG. 1 is changed to that of step S400 in FIG.
The copy instruction (write instruction) WD to the secondary side is transmitted as a set of three write instructions WD1, WD2, and WD3, and the delivery to the secondary side is shortened to one time, whereby the embodiment shown in FIG. It is possible to further increase the processing speed. In FIG. 2, steps indicating the same processes as those described in FIG. 1 are denoted by the same step numbers as in FIG.

[0029]

As described in detail above, the DASD of the present invention
According to the data transfer method between control devices (claim 1), 1
Duplicate processing can be eliminated compared to the conventional technology, in which copy processing is performed every time write processing is performed for one write instruction. Therefore, the time required for the recovery system to copy data to DASD is greatly reduced, and processing speed is increased. it can.

Further, by transferring a plurality of write instructions collectively into one write instruction, the time required for the copy processing can be further reduced, and the processing can be further speeded up. Furthermore, by releasing the channel that has been occupied conventionally during the copy processing, the channel can execute other I / O processing, so that the performance as an apparatus can be improved. ).

As described above, according to the present invention, D
The performance of the copy function between the ASD controllers can be greatly improved.

[Brief description of the drawings]

FIG. 1 is an operation flowchart illustrating a procedure of a data transfer method between DASD controllers (a procedure of data transfer from a primary control unit to a secondary control unit) as one embodiment of the present invention.

FIG. 2 is an operation flowchart illustrating a procedure (a procedure for transferring data from a primary control unit to a secondary control unit) of a modification of the data transfer method between DASD controllers according to an embodiment of the present invention.

FIG. 3 is a block diagram showing a general configuration of a system having a copy function (remote transfer function).

FIG. 4 is a diagram illustrating an example of a write instruction (CCW) issued by a host.

FIG. 5 is an operation flowchart for describing a series of flows of data transmission / reception (conventional copy processing) between an application site and a recovery site.

[Explanation of symbols]

1A, 2A Host 2A Primary control unit (DASD controller) 2B Secondary control unit (DASD controller) 3A Primary device (DASD) 3B Secondary device (DASD) 4 Interface 10A, 10B Channels 21A, 21B PRA 22A, 22B CA

Claims (3)

[Claims] In order to duplicate data held in a DASD of a normal operation system by a DASD of a recovery system, DASD controllers controlling the DASDs in these systems are connected to each other, A data transfer method for transferring update data from the normal operation system to the recovery system and copying the updated data to the DASD of the recovery system, the method comprising:
When a plurality of write instructions that cause a copy to the SD are successively issued from the higher-level device to the DASD control device, the write processing for the DASD of the normal operation system is continuously performed according to the plurality of write instructions. A method of transferring data between DASD controllers, comprising: transferring the plurality of write instructions to the recovery system after the execution, and performing a copy process on the DASD of the recovery system. 2. The data transfer method between DASD controllers according to claim 1, wherein said plurality of write instructions are combined into one write instruction and transferred to said recovery system. 3. During the execution of a copy process for DASD in the recovery system, a channel used for transmitting the plurality of write commands from the higher-level device to the DASD controller in the normal operation system is released. DAS according to claim 1 or 2, characterized in that:
Data transfer method between D control devices.