Maintaining duplex-paired storage devices during gap processing using of a dual copy function

• US 5,051,887 A
• Filed: 11/19/1990
• Issued: 09/24/1991
• Est. Priority Date: 08/25/1987
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Smith, Bret P., Berger, Blaine H.
• Primary Examiner(s): Lee, Thomas C.
• Assistant Examiner(s): Harrell, Robert B.
• Application Number:
  US07/614,983
• Time in Patent Office: 309 Days
• Field of Search: 364/200, 364/900, 371/10.1, 371/10.2, 371/38, 371/68
• US Class Current: 711/162
• CPC Class Codes: G06F 11/006 Identification G06F11/2289 takes precedence
• CPC Class Codes: G06F 11/1451 by selection of backup contents
• CPC Class Codes: G06F 11/2082 Data synchronisation
• CPC Class Codes: G06F 11/2087 with a common controller
• CPC Class Codes: G06F 13/122 where hardware performs an I/O function other than control of data transfer
• CPC Class Codes: G11C 29/74 using duplex memories, i.e. using dual copies

Abstract

A technique is described for managing data storage devices and the data stored on them. A pair of data storage devices are maintained as identical copies. The dual copy function is provided without changing any current system or application software on the host. A set of status information is maintained for each record on the data storage devices. This status information includes information which allows only the changed records to be copied to a secondary device of the dual copy. This technique also provides high availability of data storage devices by swapping a failing data storage device with a good data storage device without host intervention. A joint array structure is used for routing host requests for information to and from the correct data storage device and keeping status information concerning the data storage devices.

Claims

1. In a data processing system of the type comprising a host computer having a plurality of input/output channels, a direct access storage device controller connected to input/output channels of said host computer, and two strings of direct access storage devices connected to said direct access storage device controller, said direct access storage device controller having a memory and channel switches and storage paths, said channel switches and storage paths establishing paths between said connected input/output channels and said direct access storage devices, each direct access storage device in each of said strings having rotating recording media characterized by having a plurality of tracks on which are recorded a plurality of records, each record containing a count field and a plurality of additional fields, said fields being separated by small gaps, each record being identified by a cell number which designates an exact location of the count field for that record on a track, said direct access storage devices further including a head positioned by an actuator responsive to said direct access storage device controller for reading and writing data of said records in response to requests from said host computer, the improvement in said direct access storage device control whereby duplex-paired storage devices are maintained during gap processing by a dual copy function performed by said direct access storage device controller transparently to said host computer, said direct access storage device further comprising:

means for selecting two of said direct access storage devices as said duplex-paired storage devices, a first one of said direct access storage devices being designated as a primary storage device and a second one of said direct access storage devices being designated as a secondary storage device, said primary and secondary storage devices being physically separated but substantially identical in configuration and type;

means for synchronizing said secondary storage device with said primary storage device;

a plurality of bits of each of said cell numbers of records on said primary storage device being designated to indicate a status of a corresponding one of said count field and said plurality of fields which comprise each of said records as modified or unmodified as a result of a write operation to a record on said primary storage device;

means for setting the corresponding one of said plurality of bits to indicate a modified status when one of said count field or said plurality of fields which comprise a record is modified by a write operation to said record on said primary storage device;

means for storing modified records in said memory;

said head accessing the records on said primary storage device in sequence and means responsive to records accessed by said head for searching for a count field for each of said records;

when a count field for a record is encountered by said means for searching, said means for searching further saving in said memory the cell number from the count field for the record;

means for identifying by cell number saved in memory each record on said primary storage device containing a field that has been modified as indicated by one of said plurality of bits having been set, said step of identifying being performed during time periods corresponding to said gaps between fields; and

means for transferring from said memory to said secondary storage device only those fields updated on said primary storage device in said identifying step thereby maintaining on said secondary storage device an exact duplicate of records stored on said first storage device without degrading the performance of said data processing system.

2. The improvement in said direct access storage device as recited in claim 1 wherein said memory includes a cache, said modified records being stored in said cache by said means for storing and transferred from said cache to said secondary storage device by said transferring means.

3. The improvement in said direct access storage device controller recited in claim 2 wherein said memory further includes nonvolatile storage for storing information necessary for re-synchronizing said primary and secondary storage devices in the event of a failure of said duplex-paired storage devices, said means for synchronizing accessing said nonvolatile storage during a re-synchronization process.

4. In a data processing system of the type comprising a host computer having a plurality of input/output channels, a direct access storage device controller connected to input/output channels of said host computer, and two strings of direct access storage devices connected to said direct access storage device controller, said direct access storage device controller having a memory and channel switches and storage paths, said channel switches and storage paths establishing paths between said connected input/output channels and said direct access storage devices, each direct access storage device in each of said strings having rotating recording media characterized by having a plurality of tracks on which are recorded a plurality of records, each record containing a count field and a plurality of additional fields, said fields being separated by small gaps, each record being identified by a cell number which designates an exact location of the count field for that record on a track, said direct access storage devices further including a head positioned by an actuator responsive to said direct access storage device controller for reading and writing data in said records in response to requests from said host computer, a method of maintaining duplex-paired storage devices during gap processing by a dual copy function performed by said direct access storage device controller transparently to said host computer, said method comprising the steps of:

selecting two of said direct access storage devices as said duplex-paired storage devices, a first one of said direct access storage devices being designated as a primary storage device and a second one of said direct access storage devices being designated as a secondary storage device, said primary and secondary storage devices being physically separated but substantially identical in configuration and type;

synchronizing said secondary storage device with said primary storage device;

a plurality of bits of each of said cell numbers of records on said primary storage device being designated to indicate a status of a corresponding one of said count field and said plurality of fields which comprise each of said records as modified or unmodified as a result of a write operation to a record on said primary storage device;

setting the corresponding one of said plurality of bits to indicate a modified status when one of said count field or said plurality of fields which comprise a record is modified by a write operation to said record on said primary storage device;

storing modified records in said memory;

accessing the records on said primary storage device in sequence and searching for a count field for each of said records;

when a count field for a record is encountered, saving in said memory the cell number from the count field for the record;

identifying by cell number saved in memory each record on said primary storage device containing a field that has been modified as indicated by one of said plurality of bits having been set, said step of identifying being performed during time periods corresponding to said gaps between fields; and

transferring from said memory to said secondary storage device only those fields updated on said primary storage device in said identifying step thereby maintaining on said secondary storage device an exact duplicate of records stored on said first storage device without degrading the performance of said data processing system.

5. The method recited in claim 4 further comprising the steps of:

responding to a read request from said host computer by attempting to read data from said primary storage device;

but

if a failure in reading data from said primary storage device is detected, then reading corresponding data from said secondary storage device.

6. The method recited in claim 4 further comprising the steps of:

detecting a failure in either said primary or secondary storage devices;

selecting another one of said plurality of direct access storage devices for said duplex-paired storage devices; and

substituting the selected one of said plurality of direct access storage devices for the failed one of said primary or secondary storage devices.

7. The method recited in claim 4 wherein the step of selecting is performed by selecting said primary and secondary storage devices to be on different ones of said two strings of direct access storage devices.