Concurrent switch to shadowed device for storage controller and device errors

• US 5,870,537 A
• Filed: 03/13/1996
• Issued: 02/09/1999
• Est. Priority Date: 03/13/1996
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Yudenfriend, Harry Morris, Kern, Robert Frederic, Paulsen, Michael Aloysius, Shepard, William Chambers
• Primary Examiner(s): PALYS, JOSEPH E
• Application Number:
  US08/614,588
• Time in Patent Office: 1,063 Days
• Field of Search: 395/182.03, 395/182.04, 395/182.05, 395/182.06, 395/182.07, 395/182.09, 395/182.1, 395/182.11, 211/162, 707/202, 707/204
• US Class Current: 714/6.23
• CPC Class Codes: G06F 11/2064 while ensuring consistency
• CPC Class Codes: G06F 11/2071 using a plurality of controllers
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups

Abstract

A disaster recovery system providing remote data shadowing between a primary and a secondary site uses a method and apparatus for swapping, or switching, host directed I/O operations from a primary data storage device to a secondary data storage device in a remote copy duplex pair. Application programs running in a host processor at the primary site first quiesce all I/O operations and record updates targeted to the primary data storage device. The remote copy duplex pair is checked to verify that the secondary data storage device is a fully synchronized copy of the primary data storage device. The secondary data storage device is then swapped with the primary data storage device by terminating the remote copy duplex pair, establishing an opposite direction remote copy duplex pair such that the secondary data storage device is a primary device of the duplex pair and the primary data storage device is a shadowing device, and updating the application programs running in the primary host processor with a device address of the secondary data storage device substituted as a device address of the primary data storage device. Finally, the primary host processor resumes its application programs such that subsequent I/O operations and record updates targeted for the primary data storage device are directed through a secondary storage controller to the secondary data storage device, and copied to the primary data storage device.

Claims

1. In a data processing system providing remote data shadowing for disaster recovery purposes, said data processing system including a primary site and a secondary site, said primary site having a primary host processor running applications generating I/O operations and record updates, a primary data storage device for receiving said I/O operations and storing said record updates, and a primary storage controller coupled between said primary host processor and said primary data storage device for directing said I/O operations and said record updates to said primary data storage device, said data processing system further including a secondary site having a secondary host processor communicating with said primary host processor, a secondary data storage device for storing a copy of said record updates for data shadowing of said primary data storage device, and a secondary storage controller coupled between said secondary host processor and said secondary data storage device, said secondary storage controller further coupled to said primary host processor, a method in said primary site for switching said secondary data storage device for said primary data storage device to receive said I/O operations and said record updates from said primary host processor comprising steps of:

(a) quiescing said I/O operations and said record updates generated by said applications running in said primary host processor and to be received by said primary data storage device;

(b) verifying that said primary data storage device and said secondary data storage device are a remote copy duplex pair;

(c) terminating said remote copy duplex pair;

(d) establishing an opposite direction remote copy duplex pair such that said secondary data storage device serves as a primary device of said remote copy duplex pair and said primary data storage device serves as a shadowing device of said remote copy duplex pair;

(e) updating said applications running in said primary host processor with a device address of said secondary data storage device substituted as a device address of said primary data storage device; and

(f) resuming said remote data shadowing using said opposite direction remote copy duplex pair such that subsequent record updates generated by said applications running in said primary host processor are directed through said secondary storage controller, stored in said secondary data storage device, and copied to said primary data storage device.

2. The method of claim 1 wherein the step (a) is in response to a I/O Stop command issued to said primary host processor from an automated operations process activated at said primary site.

3. The method of claim 1 wherein the step (a) is initiated by an Error Recovery Program (ERP) running in said primary host processor responding to a error signal received from said primary storage controller indicating a permanent error was detected in said primary storage controller or said primary data storage device.

4. The method of claim 1 wherein the step (a) is initiated by said primary storage controller raising an attention interrupt to said primary host processor indicating said primary storage controller has detected a permanent error on said primary data storage device.

5. The method of claim 4 further comprising:

(g) suspending said opposite direction remote copy duplex pair when said primary storage device fails due to a permanent error to allow said subsequent record updates generated by said applications running in said primary host processor to be stored in said secondary data storage device while said primary data storage device is recovering from said permanent error.

6. The method of claim 5 wherein the step (g) further comprises change recording for said secondary data storage device such that said subsequent record updates transferred to said secondary data storage device are recorded within said secondary storage controller to allow said subsequent record updates to be copied to said primary data storage device after said primary data storage device recovers from said permanent error and said opposite direction remote copy duplex pair is reestablished.

7. The method of claim 1 wherein the step (e) further comprises setting an I/O control block within said applications running in said primary host processor to indicate said secondary data storage device is said primary device of said opposite direction remote copy duplex pair.

8. The method of claim 1 further comprising steps of:

(g) determining a set of device characteristics including a device type, model, and data format for said primary data storage device and said secondary data storage device; and

(h) verifying that said set of device characteristics indicates that said primary data storage device and said secondary data storage device are compatible.

9. In a data processing system providing remote data shadowing for disaster recovery purposes, said data processing system including a primary site and a secondary site, said primary site having a primary host processor running applications generating I/O operations and record updates, a primary data storage device for storing said record updates, and a primary storage controller coupled between said primary host processor and said primary data storage device, said secondary site having a secondary host processor, a secondary data storage device for shadowing copies of said record updates stored on said primary data storage device, and a secondary storage controller coupled between said secondary host processor and said secondary data storage device, said secondary storage controller further coupled to said primary storage controller and said primary host processor, said primary storage controller comprising:

a memory for storing said copies of said record updates to be subsequently transferred to said secondary site for remote data shadowing;

a storage path coupled for receiving said record updates from said primary host processor, said storage path directing said record updates to said primary data storage device, and further directing said copies of said record updates to said memory; and

device swapping means coupled to said primary host processor and further coupled to said secondary storage controller for interchanging said secondary data storage device with said primary data storage device by;

issuing a first request to said primary host processor to quiesce said record updates generated by said applications running in said primary host processor;

verifying that said primary data storage device and said secondary data storage device form a remote copy duplex pair;

terminating said remote copy duplex pair;

establishing an opposite direction remote copy duplex pair such that said secondary data storage device serves as a primary device of said opposite direction remote copy duplex pair and said primary data storage device serves as a shadowing device of said opposite direction remote copy duplex pair;

substituting a device address of said secondary data storage device for a device address of said primary data storage device;

issuing a second request to said primary host processor to resume said remote data shadowing using said opposite direction remote copy duplex pair; and

directing subsequent record updates generated by said applications running in said primary host processor to said secondary site, said subsequent record updates stored in said secondary data storage device and copies of said subsequent record updates shadowed to said primary data storage device.

10. The storage controller of claim 9 wherein said device swapping means further suspends said opposite direction remote copy duplex pair when said primary data storage device fails due to a permanent error and continues directing said subsequent record updates to said secondary data storage device while said primary data storage device recovers from said permanent error.

11. The storage controller of claim 10 wherein said device swapping means enables change recording such that said subsequent record updates stored in said secondary data storage device are recorded within said secondary storage controller, thereby allowing said subsequent record updates to be copied to said primary data storage device after said primary data storage device recovers from said permanent error and said opposite direction remote copy duplex pair is reestablished.

12. The storage controller of claim 9 wherein said device swapping means further includes setting a duplex pair status control block within said memory to indicate whether said opposite direction remote copy duplex pair is established, suspended, or terminated.

13. The storage controller of claim 9 wherein said device swapping means sends a third request to said primary host processor to set an I/O control block to indicate to said applications that said secondary data storage device is said primary device of said opposite direction remote copy duplex pair.

14. The storage controller of claim 9 wherein said memory is a cache memory.

15. The storage controller of claim 9 wherein said memory is a non-volatile storage.

16. A remote data duplexing system providing disaster recovery capabilities, said remote data duplexing system having a primary site and a secondary site, said secondary site receiving record updates from said primary site for data shadowing, said secondary site located remote from said primary site, said remote data duplexing system comprising:

a primary host processor at said primary site running a plurality of applications, said plurality of applications generating I/O operations and said record updates;

a primary data storage device at said primary site for receiving said I/O operations and storing said record updates;

a primary storage controller coupled between said primary host processor and said primary data storage device for directing said I/O operations and said record updates from said primary host processor to said primary data storage device;

a secondary host processor at said secondary site coupled to said primary host processor by a data link;

a secondary data storage device at said secondary site for shadowing said record updates stored on said primary data storage device, said secondary data storage device forming a remote copy duplex pair with said primary data storage device;

a secondary storage controller coupled between said secondary host processor and said secondary data storage device, said secondary storage controller further coupled to said primary storage controller and said primary host processor via a plurality of fiber optic links, said secondary storage controller directing copies of said record updates stored on said primary data storage device to said secondary data storage device for data shadowing; and

device swapping means in said primary processor and coupled to said primary storage controller and said secondary storage controller for interchanging said secondary data storage device with said primary data storage device, said secondary storage device serving as a primary device in said remote copy duplex pair and said primary storage device serving as a shadowing device in said remote copy duplex pair, said device switching means;

quiescing said I/O operations and said record updates generated by said applications running in said host processor and to be received by said primary data storage device;

terminating said remote copy duplex pair;

establishing an opposite direction remote copy duplex pair such that said secondary data storage device serves as a primary device of said opposite direction remote copy duplex pair and said primary data storage device serves as a shadowing device of said opposite direction remote copy duplex pair;

updating said applications running in said primary host processor with a device address of said secondary data storage device substituted as a device address of said primary data storage device; and

resuming said remote data duplexing using said opposite direction remote copy duplex pair such that subsequent record updates generated by said applications running in said primary host processor are directed through said secondary storage controller, stored in said secondary data storage device, and copied to said primary data storage device.

17. The remote data duplexing system of claim 16 wherein said device swapping means suspends said opposite direction remote copy duplex pair when said primary data storage device fails due to a permanent error to allow subsequent record updates generated by said applications running in said primary host processor to be stored in said secondary data storage device while said primary data storage device is recovering from said permanent error.

18. The remote data duplexing system of claim 17 wherein said device swapping means enables change recording such that said subsequent record updates stored in said secondary data storage device are recorded within said secondary storage controller, thereby allowing said subsequent record updates to be copied to said primary data storage device after said primary data storage device recovers from said permanent error and said opposite direction remote copy duplex pair is reestablished.

19. The remote data duplexing system of claim 16 wherein said device swapping means determines a set of device characteristics including a device type, model, and data format for said primary data storage device and said secondary data storage device, said device swapping means further verifying that said set of device characteristics indicate that said primary data storage device and said secondary data storage device are compatible.

20. In a data processing system providing remote data shadowing for disaster recovery purposes, said data processing system including a primary site and a secondary site, said primary site having a primary host processor running applications generating I/O operations and record updates, a primary data storage device for receiving said I/O operations and storing said record updates, and a primary storage controller coupled between said primary host processor and said primary data storage device for directing said I/O operations and said record updates to said primary data storage device, said data processing system further including a secondary site having a secondary host processor communicating with said primary host processor, a secondary data storage device for storing a copy of said record updates for data shadowing of said primary data storage device, and a secondary storage controller coupled between said secondary host processor and said secondary data storage device, said secondary storage controller further coupled to said primary host processor and said primary storage controller, an article of manufacture for swapping said secondary data storage device for said primary data storage device to receive said I/O operations and said record updates from said primary host processors,

said article of manufacture comprising a computer-readable storage medium tangibly embodying a program of executable computer instructions which may cause said data processing system to:

quiesce said I/O operations and said record updates generated by said applications running in said primary host processor and to be received by said primary data storage device;

verify that said primary data storage device and said secondary data storage device form a remote copy duplex pair;

terminate said remote copy duplex pair;

establish an opposite direction remote copy duplex pair such that said secondary data storage device serves as a primary device of said opposite direction remote copy duplex pair and said primary data storage device serves as a shadowing device of said opposite direction remote copy duplex pair;

update said applications running in said primary host processor with a device address of said secondary data storage device substituted as a device address of said primary data storage device; and

resume said remote data shadowing using said opposite direction remote copy duplex pair such that subsequent record updates targeted for said primary data storage device are directed through said secondary storage controller, stored in said secondary data storage device, and copied to said primary data storage device.