Data mirroring using batch boundaries

• US 8,850,073 B1
• Filed: 04/30/2007
• Issued: 09/30/2014
• Est. Priority Date: 04/30/2007
• Status: Expired due to Fees

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Wilkes, John
• Primary Examiner(s): Towfighi, Afshawn
• Application Number:
  US11/742,549
• Time in Patent Office: 2,710 Days
• Field of Search: 709238-244, 709246-248, 707200-206
• US Class Current: 709/248
• CPC Class Codes: G06F 11/2064 while ensuring consistency
• CPC Class Codes: G06F 11/2066 Optimisation of the communication load
• CPC Class Codes: G06F 11/2074 Asynchronous techniques
• CPC Class Codes: G06F 2201/835 Timestamp

Abstract

Provided are, among other things, systems, methods and techniques for controlling data synchronization. Local data-writing operations, made across a number of logical storage units on different data-storage devices, are divided into batches, the batches corresponding to different time segments. The data-writing operations, together with an indication of boundaries between the batches, are transmitted to a remote data-storage location. The boundaries between the batches are determined by at least one of: (1) identifying a period of time of sufficient duration during which no data-writing operations were performed and declaring a point during said period of time as one of the boundaries, (2) obtaining agreement from a plurality of entities, and (3) declaring the boundaries at pre-scheduled points in time.

Claims

1. A method of data synchronization, comprising:

dividing, by at least one computer, data-writing operations, made across a plurality of logical storage units on different data-storage devices, into batches, the batches corresponding to different time segments; and

transmitting, by the at least one computer, a plurality of the data-writing operations, together with an indication of boundaries between the batches, to a remote data-storage location,

wherein a particular one of the boundaries between the batches is determined by at least one of;

(1) identifying a period of time of sufficient duration during which no data-writing operations were performed and declaring a point during said period of time as the particular boundary, wherein identifying the period of time having the sufficient duration is to accommodate a maximum expected clock skew among a plurality of entities, and (2) obtaining agreement from the plurality of entities, wherein obtaining the agreement comprises at least some of the plurality of entities receiving a request proposing the particular boundary and the at least some of the plurality of entities responding whether or not the at least some of the plurality of entities agree with the proposed particular boundary.

2. A method according to claim 1, wherein the particular boundary is determined by obtaining the agreement from the plurality of entities, using a distributed agreement algorithm across the plurality of entities to obtain agreement from the entities in real-time.

3. A method according to claim 2, wherein in response to the request proposing the particular boundary, the at least some of the plurality of entities respond by proposing a boundary different from the particular boundary.

4. A method according to claim 2, wherein the plurality of entities comprise different data-storage device controllers.

5. A method according to claim 2, wherein the at least one computer includes plural computers, and wherein the plurality of entities comprise different ones of the plural computers.

6. A method according to claim 2, wherein the plurality of entities comprise different portions of a software program running on the at least one computer.

7. A method according to claim 1, further comprising determining that a same data block was written in an earlier first data-writing operation and in a later second data-writing operation within a same batch and, in response, suppressing transmission of the first data-writing operation.

8. A method according to claim 1, wherein data-writing operations for earlier batches are transmitted before data-writing operations for later batches.

9. The method according to claim 1, wherein the at least one computer includes plural computers, and wherein the plurality of entities comprise different ones of the plural computers, and

wherein the plural computers use different clocks, and wherein identifying the period of time having the sufficient duration comprises identifying a time gap that provides a margin to accommodate clock skew resulting from use of the different clocks.

10. A method according to claim 1, wherein the particular boundary is determined by:

identifying the period of time of sufficient duration during which no data-writing operations were performed and declaring the point during said period of time as the particular boundary; and

obtaining the agreement from the plurality of entities regarding the declared particular boundary,

wherein the declared particular boundary is used if the plurality of entities agree.

11. A method of data synchronization, comprising:

dividing, by at least one computer, data-writing operations into batches, the batches corresponding to different time segments, by obtaining real-time agreement among a plurality of automated agents that are at least partially autonomous regarding boundaries between the batches, wherein obtaining the real-time agreement comprises sending requests proposing respective boundaries, and receiving responses from at least some of the automated agents regarding whether or not the at least some of the automated agents agree with the proposed boundaries; and

transmitting, by the at least one computer, a plurality of the data-writing operations, together with an indication of boundaries between the batches, to a remote data-storage location.

12. A method according to claim 11, wherein the at least one computer includes plural computers, and wherein the automated agents comprise different ones of the plural computers.

13. A method according to claim 11, wherein the automated agents comprise different data-storage device controllers.

14. A method according to claim 11, wherein the automated agents comprise at least one of:

different application programs and different portions of a single application program.

15. A method according to claim 11, wherein in response to one of the requests, the at least some of the automated agents respond by proposing a boundary different from the boundary proposed by the one request.

16. A method according to claim 11, further comprising determining that a same data block was written in an earlier first data-writing operation and in a later second data-writing operation within a same batch and, in response, suppressing transmission of the first data-writing operation.

17. A method according to claim 11, wherein data-writing operations for earlier batches are transmitted before data-writing operations for later batches.

18. A non-transitory computer-readable storage medium storing computer-executable process steps for data synchronization, said process steps comprising:

dividing data-writing operations, made across a plurality of logical storage units on different data-storage devices, into batches, the batches corresponding to different time segments; and

transmitting a plurality of the data-writing operations, together with an indication of boundaries between the batches, to a remote data-storage location,

wherein a particular one of the boundaries between the batches is determined by at least one of;

(1) identifying a period of time of sufficient duration during which no data-writing operations were performed and declaring a point during said period of time as the particular boundary, wherein identifying the period of time having the sufficient duration is to accommodate a maximum expected clock skew among a plurality of entities, and (2) obtaining agreement from the plurality of entities, wherein obtaining the agreement comprises at least some of the plurality of entities receiving a request proposing the particular boundary and the at least some of the plurality of entities responding whether or not the at least some of the plurality of entities agree with the proposed particular boundary.

19. A computer-readable medium according to claim 18, wherein the particular boundary is determined by obtaining the agreement from the plurality of entities, using a distributed agreement algorithm across the plurality of entities to obtain agreement from the entities in real-time.

20. A computer-readable medium according to claim 18, wherein in response to the request proposing the particular boundary, the at least some of the entities respond by proposing a boundary different from the particular boundary.

21. A computer-readable medium according to claim 18, wherein the plurality of entities comprise different ones of plural computers, and wherein the plural computers use different clocks, and wherein identifying the period of time having the sufficient duration comprises identifying a time gap that provides a margin to accommodate clock skew resulting from use of the different clocks.

22. A computer-readable medium according to claim 18, wherein the particular boundary is determined by:

identifying the period of time of sufficient duration during which no data-writing operations were performed and declaring the point during said period of time as the particular boundary; and

obtaining the agreement from the plurality of entities regarding the declared particular boundary,

wherein the declared particular boundary is used if the plurality of entities agree.