Asynchronous remote data copying using subsystem to subsystem communication

Claim 2

In a data copy asynchronous system, the data copy system including a first plurality of subsystems at a primary site, the first plurality of subsystems being interconnected by a first coupling means; a second plurality of subsystems at a site remote from the primary site, the second plurality of subsystems being interconnected by a second coupling means, each of the second plurality of subsystems being independently coupled to one of the first plurality of subsystems, an updating system for providing sequence consistent write operations comprising: in the first plurality of subsystems; means for distributing a sequence signal to the first plurality of subsystems; means responsive to the distributing means for sending a periodic synchronizing time-denominated checkpoint signal to the first plurality of subsystems, the checkpoint signal having a predetermined relationship with the sequence signal; means responsive to the checkpoint sending means for sending updated data and the checkpoint signal to each of the peer coupled second plurality of subsystems; and in the second plurality of subsystems; means for receiving the updated data and checkpoint signals; and means for coordinating a writing of the updated data; wherein the first and second plurality of subsystems form a database management system.

Claim 3

The updating system of claim 1 in which the distributing means further comprises: first means for activating each of the first plurality of subsystems to communicate with the other of the first plurality of subsystems; second means responsive to the first activating means for activating the first plurality of subsystems to communicate with its peer coupled subsystem of the second plurality of subsystems; means responsive to the second activating means for building configuration tables in each of the first plurality of subsystems such that each of the first plurality of subsystems can identify all of the other first plurality of subsystems; and means responsive to the building means for synchronizing the first plurality of subsystems.

Claim 4

The updating system of claim 1 in which the checkpoint signal sending means comprises: means for sending a checkpoint message to the first plurality of subsystems; and means responsive to the checkpoint message sending means for inserting the checkpoint message into an update data sequence of the first plurality of subsystems.

Claim 5

The updating system of claim 1 in which the coordinating means comprises: means for receiving copy active messages in the second plurality of subsystems; means responsive to the copy active sending means for building copy active tables in the second plurality of subsystems; means responsive to the building means for synchronizing copy operations of second plurality of subsystems; means responsive to the synchronizing means for receiving the checkpoint messages in the second plurality of subsystems; means responsive to the checkpoint messages receiving means for performing a rendezvous for all checkpoint messages in the second plurality of subsystems; and means responsive to the performing means for providing the updated data to second plurality of subsystems.

Claim 8

The method of claim 5 in which the distributing step (a) further comprises the steps of: activating each of the first plurality of subsystems to communicate with the second plurality of subsystems; activating the second plurality of subsystems to communicate with its peer coupled subsystem at the remote site; building configuration tables in each of the first plurality of subsystems such that each of the first plurality of subsystems can identify all of the other first plurality of subsystems at the primary site; and synchronizing the subsystems at the primary site.

Claim 9

The method of claim 5 in which the checkpoint signal sending step (b) further comprises the steps of: sending a checkpoint message to the first plurality of subsystems; and inserting the checkpoint message into an update data sequence of first plurality of subsystems.

Claim 10

The method of claim 5 in which the coordinating step (d) comprises the steps of: receiving copy active messages in the second plurality of subsystems; building copy active tables in the second plurality of subsystems; synchronizing copy operations of second plurality of subsystems; receiving the checkpoint messages in the second plurality of subsystems; performing a rendezvous for all checkpoint messages in the second plurality of subsystems; and providing the updated data to the second plurality of subsystems.

Claim 11

In a system for remote copying of data at a secondary site having a coupling plurality of storage device (DASD) subsystems that are interconnected by a first coupling means, said secondary DASD subsystems being coupled to counterpart ones of a plurality of DASD subsystems at a remote primary site, said secondary DASD subsystem being interconnected by a second coupling means said system including means at the primary site for initiating a start copy operation, a method comprising the steps of: (a) at the primary and secondary sites, forming an copyset groups of DASD subsystems respectively; (b) at the primary site, (1) causing each of the m subsystems to asynchronously generate a sequence of updated write records, (2) ordering each of said sequences by embedding common clock values and a periodic checkpoint message with a common clock value and increasing a sequence number in each of said sequences, (3) coupling counterpart DASD subsystems between the sites and asynchronously communicating sequences from each of the primary site subsystems into a buffered portion of a counterpart secondary site subsystem; and (c) at the secondary site, writing the buffered sequences to DASD in the counterpart subsystems as a function of each checkpoint message wherein data integrity is maintained in a database management system formed by the remote primary and secondary sites of DASD sub-systems.

Claim 12

In a system for remote copying of data at a secondary site having a coupling plurality of storage device (DASD) subsystems that are interconnected by a first coupling means, said secondary DASD subsystems being coupled to counterpart ones of a plurality of DASD subsystems at a remote primary site, said secondary DASD subsystem being interconnected by a second coupling means said system including means at the primary site for initiating a start copy operation, a method comprising the steps of: (a) at the primary and secondary sites, forming an copyset groups of DASD subsystems respectively; (b) at the primary site, (1) causing each of the m subsystems to asynchronously generate a sequence of updated write records, (2) ordering each of said sequences by embedding common clock values and a periodic checkpoint message with a common clock value and increasing a sequence number in each of said sequences, (3) coupling counterpart DASD subsystems between the sites and asynchronously communicating sequences from each of the primary site subsystems into a buffered portion of a counterpart secondary site subsystem; and (c) at the secondary site, writing the buffered sequences to DASD in the counterpart subsystems as a function of each checkpoint message wherein data integrity is maintained in a database management system formed by the remote primary and secondary sites of DASD sub-systems.

Claim 13

In a system for remote copying of data at a secondary site, a CPU attached plurality of direct access storage device (DASD) subsystems, said secondary DASD subsystems being coupled to counterpart ones of a plurality of DASD subsystems CPU attached at a remote primary site, said system including means at the primary site for initiating a start copy operation, a method comprising the steps of: (a) at the primary site responsive to initiation of the start copy operation, ascertaining at each subsystem the subset of the plurality of DASD subsystems forming the copyset group, designating one of the plurality of subsystems as a clocking and checkpoint message source, each checkpoint message including a sequence clock value and an increased sequence number; (b) at the secondary site, repeating step (a) for counterpart ones of the DASD subsystems; (c) at the primary site, periodically generating clocking signals and checkpoint messages by said designated subsystem and broadcasting said signals and messages as they occur to other subsystems forming the copyset group including itself, at each subsystem in the copyset group: (1) asynchronously forming a local sequence of updated records, (2) embedding slid signals and messages into the sequence to form a time discriminated total ordering of updated records, and (3) remitting at least a portion of the sequence to a buffer portion of the counterpart DASD subsystem at the secondary site; and (d) at the secondary site, applying a checkpoint message to the designated subsystem operative as a synchronizing source by each subsystem in the copyset group and writing the sequences or portions thereof stored in the buffered portions of said subsystems to DASD only a signal from said designated subsystem indicative of its receipt of all send messages wherein data integrity is maintained in a database management system formed by the remote primary and secondary sites of DASD sub-systems.

Claim 14

In a system for remote copying of data at a secondary site, a CPU attached plurality of direct access storage device (DASD) subsystems, said secondary DASD subsystems being coupled to counterpart ones of a plurality of DASD subsystems CPU attached at a remote primary site, said system including means at the primary site for initiating a start copy operation, a method comprising the steps of: (a) at the primary site responsive to initiation of the start copy operation, ascertaining at each subsystem the subset of the plurality of DASD subsystems forming the copyset group, designating one of the plurality of subsystems as a clocking and checkpoint message source, each checkpoint message including a sequence clock value and an increased sequence number; (b) at the secondary site, repeating step (a) for counterpart ones of the DASD subsystems; (c) at the primary site, periodically generating clocking signals and checkpoint messages by said designated subsystem and broadcasting said signals and messages as they occur to other subsystems forming the copyset group including itself, at each subsystem in the copyset group: (1) asynchronously forming a local sequence of updated records, (2) embedding slid signals and messages into the sequence to form a time discriminated total ordering of updated records, and (3) remitting at least a portion of the sequence to a buffer portion of the counterpart DASD subsystem at the secondary site; and (d) at the secondary site, applying a checkpoint message to the designated subsystem operative as a synchronizing source by each subsystem in the copyset group and writing the sequences or portions thereof stored in the buffered portions of said subsystems to DASD only a signal from said designated subsystem indicative of its receipt of all send messages wherein data integrity is maintained in a database management system formed by the remote primary and secondary sites of DASD sub-systems.

Claim 15

The method according to claim 10, wherein each message signifies that all DASD records with a clock signal that have a sequence number that is less than the checkpoint message sequence clock value have been transmitted to the counterpart secondary subsystem.

Claim 16

In a data copy asynchronous system, the data copy system including a first plurality of subsystems at a primary site, the first plurality of subsystems being interconnected by a first coupling means; a second plurality of subsystems at a site remote from the primary site, the second plurality of subsystems being interconnected by a second coupling means, each of the second plurality of subsystems being independently coupled to one of the first plurality of subsystems, a method for providing sequence consistent write operations comprising the steps of: in the first plurality of subsystems; (a) distributions a checkpoint signal to the first plurality of subsystems; (b) sending a checkpoint signal to the first plurality of subsystems, the checkpoint signal having a predetermined relationship with the sequence signal; (c) sending updated data and the checkpoint signal to each of the peer coupled second plurality of subsystems; and in the second plurality of subsystems; (d) receiving the updated data and checkpoint signals; and (e) coordinating a writing of the updated data wherein data integrity is maintained in a database management system formed by the first and second plurality of subsystems.

Claim 17

In a data copy asynchronous system, the data copy system including a first plurality of subsystems at a primary site, the first plurality of subsystems being interconnected by a first coupling means; a second plurality of subsystems at a site remote from the primary site, the second plurality of subsystems being interconnected by a second coupling means, each of the second plurality of subsystems being independently coupled to one of the first plurality of subsystems, a method for providing sequence consistent write operations comprising the steps of: in the first plurality of subsystems; (a) distributions a checkpoint signal to the first plurality of subsystems; (b) sending a checkpoint signal to the first plurality of subsystems, the checkpoint signal having a predetermined relationship with the sequence signal; (c) sending updated data and the checkpoint signal to each of the peer coupled second plurality of subsystems; and in the second plurality of subsystems; (d) receiving the updated data and checkpoint signals; and (e) coordinating a writing of the updated data wherein data integrity is maintained in a database management system formed by the first and second plurality of subsystems.

Claim 18

The method of claim 12 in which the distributing step further comprises the steps of: activating each of the first plurality of subsystems to communicate with the other of the first plurality of subsystems; activating the first plurality of subsystems to communicate with its peer coupled subsystem of the second plurality of subsystems; building configuration tables in each of the first plurality of subsystems such that each of the first plurality of subsystems can identify all of the other first plurality of subsystems; and synchronizing the first plurality subsystems.

Claim 19

The method of claim 12 in which the checkpoint signal sending step (c) further comprises the steps of: sending a checkpoint message to the first plurality of subsystems; and inserting the checkpoint message into an update data sequence of first plurality of subsystems.

Claim 20

The method of claim 12 in which the coordinating step (e) further comprises the steps of: receiving copy active messages in the second plurality of subsystems; building copy active tables in the second plurality of subsystems; synchronizing copy operations of second plurality of subsystems; receiving the checkpoint messages in the second plurality of subsystems; performing a rendezvous for all checkpoint messages in the second plurality of subsystems; and providing the updated data to second plurality of subsystems.

Claim 21

A computer readable medium containing program instructions for providing sequence consistent write operations in an updating system of a data copy asynchronous system having a first plurality of interconnected primary subsystems and a second plurality of interconnected remote subsystems, each of the second plurality of subsystems coupled to one of the first plurality of subsystems, the program instructions comprising: distributing a sequence signal to the first plurality of subsystems; sending a checkpoint signal to the first plurality of subsystems in response to the sequence signal, the checkpoint signal and the sequence signal having a predetermined relationship; sending updated data and the checkpoint signal to each corresponding second plurality of subsystems; receiving the updated data and checkpoint signals; and coordinating a writing of the updated data wherein integrity is maintained in a database management system formed by the first and second plurality of sub-systems.

Claim 22

A computer readable medium containing program instructions for providing sequence consistent write operations in an updating system of a data copy asynchronous system having a first plurality of interconnected primary subsystems and a second plurality of interconnected remote subsystems, each of the second plurality of subsystems coupled to one of the first plurality of subsystems, the program instructions comprising: distributing a sequence signal to the first plurality of subsystems; sending a checkpoint signal to the first plurality of subsystems in response to the sequence signal, the checkpoint signal and the sequence signal having a predetermined relationship; sending updated data and the checkpoint signal to each corresponding second plurality of subsystems; receiving the updated data and checkpoint signals; and coordinating a writing of the updated data wherein integrity is maintained in a database management system formed by the first and second plurality of sub-systems.