COMPUTER ARCHITECTURE AND METHOD OF OPERATION FOR MULTI-COMPUTER DISTRIBUTED PROCESSING WITH SYNCHRONIZATION

• US 20090235034A1
• Filed: 12/23/2008
• Published: 09/17/2009
• Est. Priority Date: 04/21/2005
• Status: Active Grant

• Current Assignee: Waratek Limited (Waratek Holdings Ltd.)
• Original Assignee: John Matthew Holt
• Inventors: Holt, John Matthew
• Granted Patent: US 7,818,296 B2
• Time in Patent Office: Days
• US Class Current: 711/150
• CPC Class Codes: G06F 15/16 Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs coordinating program control therefor G06F9/52; in regulating and control system G05B
• CPC Class Codes: G06F 8/456 Parallelism detection

Abstract

The present invention discloses a modified computer architecture (50, 71, 72) which enables an applications program (50) to be run simultaneously on a plurality of computers (M1, . . . Mn). Shared memory at each computer is updated with amendments and/or overwrites so that all memory read requests are satisfied locally. During initial program loading (75), or similar, instructions which result in the application program (50) acquiring (or releasing) a lock on a particular asset (50A, 50X-50Y) (synchronization) are identified. Additional instructions are inserted (162, 163) to result in a modified synchronization routine with which all computers are updated.

Claims

1. A single computer configured to operate in a multiple computer system comprising a plurality of computers interconnected by a communications network without forming a distributed shared memory arrangement, said single computer having at least one application program each written to operate on only one computer but running simultaneously on said plurality of computers;

wherein different portions of said application program(s) execute substantially simultaneously on different ones of said computers and for each portion a like plurality of substantially identical memory locations are created;

each in the corresponding computer and each having a substantially identical name; and

said single computer including a lock means applicable to all said computers wherein any computer wishing to exclusively utilize a named memory location therein acquires an authorizing lock from said lock means which permits said utilization and which prevents all the other computers from exclusively utilizing their corresponding named memory location until said authorizing lock is relinquished.

2. The single computer as claimed in claim 1 wherein said lock means includes an acquire lock routine and a release lock routine, and both said routines are included in modifications made to said application program.

3. The single computer as claimed in claim 2 wherein said lock means further includes a shared table listing said named memory locations in use by any said computer, a lock status for each said memory location, and a queue of any pending lock acquisitions.

4. The single computer as claimed in claim 3 and connected to an additional computer not running said application program and said additional computer being connected to said communications network.

5. The single computer as claimed in claim 2 wherein each said application program is modified before, during, or after loading by inserting said acquire lock routine and said release lock routine to modify each instance at which said application program acquires and releases respectively a lock on a memory location.

6. The single computer as claimed in claim 5 wherein the application program is modified in accordance with a procedure selected from the group of procedures consisting of re-compilation at loading, pre-compilation prior to loading, compilation prior to loading, just-in-time compilation, and re-compilation after loading and before execution of the relevant portion of application program.

7. The single computer as claimed in claim 2 wherein said modified application program is transferred to all said computers in accordance with a procedure selected from the group consisting of master/slave transfer, branched transfer and cascaded transfer.

35. A single computer as in claim 1, wherein each said computer in operating its application program portion utilizes a memory location only in local independent memory physically located in each said computer.

8. A single computer arranged to operate within a plurality of computers interconnected via a communications link without forming a distributed shared memory arrangement, said plurality of computers operating substantially simultaneously at least one application program each written to operate on only one computer;

wherein each said computer substantially simultaneously executes a different portion of said application program(s);

each said computer in operating its application program portion utilizes a memory location only in local independent memory physically located in each said computer;

the contents of the local independent memory utilized by each said computer is fundamentally similar but not, at each instant, identical; and

every one of said computers has an acquire lock routine and a release lock routine which permit exclusive utilization of the local memory location only by said single computer and each of the remainder of said plurality of computers is locked out of exclusive utilization of their corresponding memory location.

9. The single computer as claimed in claim 8 wherein the local memory capacity allocated to each said application program is substantially identical and the total memory capacity available to each said application program is said allocated memory capacity.

10. The single computer as claimed in claim 8 wherein said communications link communicates at a data transfer rate which is substantially less than the local memory read rate.

11. The single computer as claimed in claim 8 wherein at least some of said plurality of computers are manufactured by different manufacturers and/or have different operating systems.

12. A method of running on a single computer at least one application program each written to operate only on one computer;

said single computer being configured to operate with a plurality of computers which are interconnected by means of a communications network without forming a distributed shared memory arrangement;

said method comprising the steps of;

(i) executing different portions of said application program(s) substantially simultaneously on different ones of said computers and for each said portion creating a like plurality of substantially identical memory locations each in the corresponding computer and each having a substantially identical name; and

(ii) requiring any of said computers wishing to exclusively utilize a named memory location therein to acquire an authorizing lock which permits said utilization and which prevents all the other computers from exclusively utilizing their corresponding named memory location until said authorizing lock is relinquished.

13. A method as claimed in claim 12 including the further step of:

(iii) providing each said computer with a distributed run time means to communicate between said computers via said communications network.

14. A method as claimed in claim 13 including the further step of:

(iv) providing a shared table accessible by each said distributed run time means and in which is stored the identity of any computer which currently has to access a memory location, together with the identity of the memory location.

15. A method as claimed in claim 14 including the further step of:

(v) associating a counter means with said shared table, said counter means storing a count of the number of said computers which seek access to said memory location.

16. A method as claimed in claim 15 including the further step of:

(vi) providing an additional computer on which said shared program does not run and which hosts said shared table and counter, said additional computer being connected to said communications network.

29. A computer program product comprising a set of program instructions stored in a storage medium and operable to permit a single computer or a plurality of computers to carry out the method as claimed in claim 12.

30. A single computer configured to operate with a plurality of computers interconnected via a communication network and operable to ensure consistent synchronization of an application program written to operate only on one computer but running substantially simultaneously on all said computers, said single computer being programmed to carry out the method as claimed in claim 12.

33. A single computer configured to operate with a plurality of computers interconnected via a communication network and operable to ensure consistent synchronization of an application program written to operate only on one computer but running substantially simultaneously on all said computers, said single computer being loaded with the computer program product as claimed in claim 29.

36. A method as in claim 12, wherein each said computer in operating its application program portion utilizes a memory location only in local independent memory physically located in each said computer.

17. A method of ensuring for a single computer consistent synchronization of an application program written to operate only on one computer but different portions of which are to be executed substantially simultaneously each on a different one of a plurality of computers;

said plurality of computers including said single computer and being interconnected via a communications network without forming a distributed shared memory arrangement;

said method comprising the steps of;

(i) scrutinizing said application program at, or prior to, or after loading to detect each program step defining a synchronization routine; and

(ii) modifying said synchronization routine to ensure exclusive utilization of a memory location by only one of said plurality of computers and preventing all the remaining computers of said plurality of computers from substantially simultaneously exclusively utilizing their corresponding memory locations.

18. The method claimed in claim 17 wherein step (ii) comprises the steps of:

(iii) loading and executing said synchronization routine on said single computer, (iv) modifying said synchronization routine by said single computer, and (v) transferring said modified synchronization routine to each of the remaining computers.

19. The method as claimed in claim 18 wherein said modified synchronization routine is supplied by said single computer direct to each of said remaining computers.

20. The method as claimed in claim 18 wherein said modified synchronization routine is supplied in cascade fashion from said single computer sequentially to each of said remaining computers.

21. The method claimed in claim 17 wherein step (ii) comprises the steps of:

(vi) loading and modifying said synchronization routine on said single computer, (vii) said single computer sending said unmodified synchronization routine to each of the remaining computers, and (viii) each of said remaining computers modifying said synchronization routine after receipt of same.

22. The method claimed in claim 21 wherein said unmodified synchronization routine is supplied by said single computer directly to each of said remaining computers.

23. The method claimed in claim 21 wherein said unmodified synchronization routine is supplied in cascade fashion from said single computer sequentially to each of said remaining computers.

24. The method as claimed in claim 17 including the further step of:

(ix) modifying said application program utilizing a procedure selected from the group of procedures consisting of re-compilation at loading, pre-compilation prior to loading, compilation prior to loading, just-in-time compilation, and re-compilation after loading and before execution of the relevant portion of application program.

25. The method as claimed in claim 17 including the further step of:

(x) transferring the modified application program to all said computers utilizing a procedure selected from the group consisting of master/slave transfer, branched transfer and cascaded transfer.

31. A computer program product comprising a set of program instructions stored in a storage medium and operable to permit a single computer or a plurality of computers to carry out the method as claimed in claim 17.

32. A single computer configured to operate with a plurality of computers interconnected via a communication network and operable to ensure consistent synchronization of an application program written to operate only on one computer but running substantially simultaneously on all said computers, said single computer being programmed to carry out the method as claimed in claim 17.

34. A single computer configured to operate with a plurality of computers interconnected via a communication network and operable to ensure consistent synchronization of an application program written to operate only on one computer but running substantially simultaneously on all said computers, said single computer being loaded with the computer program product as claimed in claim 31.

37. A method as in claim 17, wherein each said computer in operating its application program portion utilizes a memory location only in local independent memory physically located in each said computer.

26. In a multiple thread processing computer operation taking place on a single computer configured to operate with a plurality of computers and in which individual threads of a single application program written to operate only on said single computer are simultaneously being processed each on a corresponding different one of said plurality of computers which are interconnected via a communications link without forming a distributed shared memory arrangement;

and in which memory locations in local independent memory physically associated with the computer processing each thread have corresponding memory locations in the local independent memory of each other said computer;

the improvement comprising permitting only said single computer to exclusively utilize a memory location and preventing all the remaining computers from substantially simultaneously exclusively utilizing their corresponding memory location.

27. The improvement as claimed in claim 26 wherein a memory location residing in the memory associated with one said thread and to be utilized has its identity communicated by the computer of said one thread to a shared table accessible by all other said computers.

28. The improvement as claimed in claim 26 wherein a memory location residing in the memory associated with one said thread and to be utilized has its identity transmitted to the computer associated with another said thread and is transmitted thereby to a shared table accessible by all said other computers.

38. In a multiple computer system including a plurality of single computers interconnected via a communications link, a method of operating an application program having code written to include a plurality of threads all intended to execute on and reference a single independent computer having a single central processing unit (CPU) or symmetric multiple processing units and a single independent local memory that is not shared with any other computer of said plurality of single computers, said method comprising:

(a) allocating said plurality of application program code threads amongst said plurality of interconnected single computers so that at least a portion of the application program including at least one first thread of said application program code is allocated a first particular one of said plurality of single computers for execution thereon and at least one second thread of said application program code is allocated a second particular one of said plurality of single computers for execution thereon;

(b) operating each single computer so that the memory locations of its single independent local memory are only addressable by the thread or threads allocated for execution thereon;

(c) substantially simultaneously executing said allocated application program code thread or threads by those single computers having an application program code thread or threads allocated thereto for execution;

(d) storing, during or after execution, by said single computer in its single independent local memory, an original version of each application memory value addressable by execution of said thread or threads allocated thereto;

(e) communicating an update via said communications link, a replica version of each new original version of each application memory value which has overwritten a previous corresponding application memory value as a consequence of execution of said thread or threads on any said single computer, to the single independent local memory of all other ones of said plurality of single computers; and

(f) storing in said single independent local memory of all of said plurality of single computers said communicated updated replica version of each and every application memory value;

(g) locking memory locations in each said single computer wherein any single computer wishing to utilize a memory location therein acquires an authorizing lock within the multiple computer system which permits said utilization of said memory location and which prevents all the other of the single computers from utilizing their corresponding memory location until said authorizing lock is relinquished; and

(h) the stored application memory values, including said original versions and said replica versions, of all said single computers thereby remaining, subject to a communications link updating transmission delay, substantially identical without any of said single computers addressing the independent local memory of any other one of said plurality of single computers during execution of their allocated thread or threads.

40. The method as claimed in claim 38, wherein the communicating includes executing an updating propagation routine.

41. The method as claimed in claim 40, wherein executing the updating propagation routine includes communicating the identity of updated, changed, or manipulated memory location, and the updated, manipulated or changed value(s) or content(s) of the manipulated memory location.

42. The method as claimed in claim 41, wherein the updating communication is performed via a slow network communications link or path.

43. The method as claimed in claim 40, wherein each of said plurality of single computers receiving an update communication from the network, writes the memory location value(s) received to the local independent memory location corresponding to the identified memory location received.

44. The method as claimed in claim 42, wherein the communications link includes a communications link over the Internet.

45. The method as claimed in claim 42, wherein the slow network is a network having a communications speed of operation that is an order of magnitude slower than the speed of operation of the bus on the single computers.

46. The method as claimed in claim 40, wherein the communications between the single computers although routed through the single computer machine hardware are controlled by an individual distributed runtime (DRT) within each single computer, the distributed runtime coordinating particular communications between the plurality of single computers.

47. The method as claimed in claim 46, wherein the DRT coordinates communications between the single computers that is transport, protocol, and communications link independent.

48. The method as claimed in claim 40, wherein:

all reading of memory locations or data is satisfied locally from the single computer local independent memory because a current value of all (or some subset of all) memory locations is stored on the single computer carrying out the processing of the code threads which generates the demand to read from memory; and

all writing of memory locations or data is satisfied locally from the single computer local independent memory because a current value of all (or some subset of all) memory locations is stored on the single computer carrying out the processing of the code threads which generates the demand to write to memory;

wherein the demand to write or rewrite memory being relatively lower than the demand to read from memory, so that memory locations can be continually updated at a relatively low speed via a slow and inexpensive commodity network, yet this sufficient to meet the application program demand for writing to memory.

39. A computer program product stored on a tangible machine-readable medium and including executable instructions for, in a multiple computer system including a plurality of single computers interconnected via a communications link, implementing a method of operating an application program having code written to include a plurality of threads all intended to execute on and reference a single independent computer having a single central processing unit (CPU) or symmetric multiple processing units and a single independent local memory that is not shared with any other computer of said plurality of single computers, said executable instruction performing a method comprising:

(a) allocating said plurality of application program code threads amongst said plurality of interconnected single computers so that at least a portion of the application program including at least one first thread of said application program code is allocated a first particular one of said plurality of single computers for execution thereon and at least one second thread of said application program code is allocated a second particular one of said plurality of single computers for execution thereon;

(b) operating each single computer so that the memory locations of its single independent local memory are only addressable by the thread or threads allocated for execution thereon;

(c) substantially simultaneously executing said allocated application program code thread or threads by those single computers having an application program code thread or threads allocated thereto for execution;

(d) storing, during or after execution, by said single computer in its single independent local memory, an original version of each application memory value addressable by execution of said thread or threads allocated thereto;

(e) communicating an update via said communications link, a replica version of each new original version of each application memory value which has overwritten a previous corresponding application memory value as a consequence of execution of said thread or threads on any said single computer, to the single independent local memory of all other ones of said plurality of single computers; and

(f) storing in said single independent local memory of all of said plurality of single computers said communicated updated replica version of each and every application memory value; and

(g) locking memory locations in each said single computer wherein any single computer wishing to utilize a memory location therein acquires an authorizing lock within the multiple computer system which permits said utilization of said memory location and which prevents all the other of the single computers from utilizing their corresponding memory location until said authorizing lock is relinquished; and

(h) the stored application memory values, including said original versions and said replica versions, of all said single computers thereby remaining, subject to a communications link updating transmission delay, substantially identical without any of said single computers addressing the independent local memory of any other one of said plurality of single computers during execution of their allocated thread or threads.

49. A multiple computer system comprising:

a plurality of single computers interconnected via a communications link;

each of said plurality of single computers;

(a) has a single central processing unit (CPU) or symmetric multiple processing units, (b) a single independent local memory which is not shared with any other computer of said plurality of single computers, and (c) that is independent of the other ones of said plurality of single computers;

each one of said plurality of single computers is adapted to store and execute at least one thread of a substantially similar copy of an application program having code written to operate only on a single computer system, the code including a plurality of threads all intended to execute in cooperation with a single memory; and

each said single independent local memory has memory locations only addressable and accessible by the thread or threads of the corresponding single computer executing the corresponding copy of said application program code, and each said independent local memory has stored therein all application memory values arising from the execution of all said threads of all said plurality of computers;

(i) means for allocating or receiving an allocation of said plurality of threads of said stored application program code amongst said plurality of single computers so that at least a portion of the application program including at least one thread of said application program code is allocated to a corresponding one of said plurality of single computers for execution thereon by one or more threads of that one of said plurality of single computers;

(ii) means for substantially simultaneously executing said application program code thread or threads by those ones of said plurality of single computers having an application program code thread or threads allocated thereto;

(iii) a memory storage for storing in said single independent local memory of each said single computer a replica of each and every application memory value addressable by each and every said thread of each and every one of said plurality of computers;

(iv) means for communicating via said communications link each new application memory value which has overwritten a previous corresponding value as a consequence of execution of said application program code on any one of said single computers, to the single independent local memory of all other ones of said plurality of single computers, so that the values of said replica application memory values of all said single computers remain, subject to an updating transmission delay, substantially identical; and

(v) lock means applicable to each said single computer wherein any single computer wishing to utilize a memory location therein acquires an authorizing lock from said lock means which permits said utilization of said memory location and which prevents all the other of said single computers from utilizing their corresponding memory location until said authorizing lock is relinquished.

50. The multiple computer system as in claim 49, wherein the communications link comprises an Internet network communications link.