Resource sharing facility functions as a controller for secondary storage device and is accessible to all computers via inter system links

• US 5,848,241 A
• Filed: 01/09/1997
• Issued: 12/08/1998
• Est. Priority Date: 01/11/1996
• Status: Expired due to Term

• Current Assignee: Hanger Solutions, LLC (IP Investments Group LLC)
• Original Assignee: Openframe Corporation Ltd.
• Inventors: Rotschild, Michael, Misinai, Tsvi, Kremer, Arie, Alazraki, Shmuel
• Primary Examiner(s): Luu, Le Hien
• Application Number:
  US08/781,151
• Time in Patent Office: 698 Days
• Field of Search: 707/205, 707/201, 395/876, 395/200.43, 395/200.47, 711/112, 711/119
• US Class Current: 709/213
• CPC Class Codes: G06F 12/084 with a shared cache
• CPC Class Codes: G06F 9/54 Interprogram communication

Abstract

A method and architecture for sharing data items using a resource sharing facility with a medium level cache accessible to all of the computers which are to share and connected to the computers by fast intersystem links. The resource sharing facility also acts as an ordinary secondary storage controller and in the resource sharing facility the control information and logic is provided which is needed for the mapping of cache data to the secondary storage and vice versa.

Claims

1. A method of sharing data items and locks and of supporting remote procedure calls and coordinated process groups between a plurality of computers each having one or more respective central processing units, respective main storage associated with the central processing units of each computer and respective high-level caches maintained in the main storage of the respective computer, said method comprising the steps of:

(a) providing a resource sharing facility with a medium level cache and a resource sharing facility database and capable of selectively storing a set of data items in said medium level cache and in said resource sharing facility database;

(b) connecting said resource sharing facility to all of said computers by intersystem links so that said resource sharing facility is accessible to all of said computers;

(c) operating said resource sharing facility as a controller for at least one secondary storage;

(d) maintaining in said resource sharing facility control information and logic needed for mapping medium level cache data to said secondary storage and said! secondary storage data to said medium level cache for each data item in said medium level cache where said resource sharing facility optionally unifies redundant control information for said computers; and

(e) synchronizing, when required, cache data maintained in the resource sharing facility and persistent data residing in said secondary storage by delegating input/output operations from said computer to said resource sharing facility.

2. The method defined in claim 1 wherein said resource sharing facility database is managed by a resource sharing facility database management system implemented directly in the resource sharing facility and operated for maintenance and use of queues needed for online transaction processing and batch job execution.

3. The method defined in claim 1, further comprising the step of supporting shared database journal files in the resource sharing facility.

4. The method defined in claim 1, further comprising the step of effecting transaction and job load balancing operations using the resource sharing facility for sharing computer independent session context data and job context data.

5. The method defined in claim 1, further comprising the step of implementing management of said resource sharing facility database by a resource sharing facility database management system in at least one of said computers.

6. The method defined in claim 5 wherein data maintained by said resource sharing facility database management system is stored in said main storage of said one of said computers and is not shared between said computers.

7. The method defined in claim 1 wherein the shared data items are selected from:

structured data of a database management system data set,

structured data of a nondatabase management system data set, and

unstructured data items.

8. The method defined in claim 7 wherein said structured data is selected from the group which consists of:

data of an MVS DB2 database,

data of an IMS database,

data of an ADABAS database,

data of a VSAM file,

data of a DB2/6000 database,

data of an Oracle database,

data of a Sybase database,

data of an Informix database,

data of an Ingres database,

data of a Microsoft®

SQL server database,

data of an SQL/DS database,

data of a CA-IDMS database,

data of a DataCom/DB database,

data of IAM files, and

data managed by any other DBMS.

9. The method defined in claim 1 wherein said computers are selected from the group which consists of IBM®

mainframes and compatible, UNIX machines, NT machines, PCs and compatibles or any other type of computer and combinations thereof.

10. The method defined in claim 1, further comprising the steps of:

reading data items from said medium-level cache with a read request; and

locking data items read from said medium-level cache against concurrent update or read in the medium-level cache, with the same read request.

11. The method defined in claim 1 wherein at least some of the computers are heterogeneous and are connected to a heterogeneous computer system secondary storage controller and wherein some shared data resides in said heterogeneous computer system secondary storage controller which is used to read and write but not lock said some shared data and where mapping between the portion of said some shared data which is kept also in the medium level cache and its permanent copy in secondary storage of the heterogeneous computer system secondary storage controller is not delegated to said resource sharing facility.

12. A method of sharing data items and locks and of supporting remote procedure calls and coordinated process groups between a plurality of computers each having one or more respective central processing units, respective main storage associated with the central processing units of each computer and respective high-level caches maintained in the main storage of the respective computer, said method comprising the steps of:

(a) providing a resource sharing facility with a medium level cache and a resource sharing facility database and capable of selectively storing a set of data items in said medium level cache and in said resource sharing facility database;

(b) connecting said resource sharing facility to all of said computers by intersystem links so that said resource sharing facility is accessible to all of said computers;

(c) operating said resource sharing facility as a controller for at least one secondary storage;

(d) maintaining in said resource sharing facility control information and logic needed for mapping medium level cache data to said secondary storage and secondary storage data to said medium level cache for each data item in said medium level cache where said resource sharing facility optionally unifies redundant control information for said computers; and

(e) synchronizing when required, cache data maintained in the resource sharing facility and persistent data residing in said secondary storage by delegating input/output operations from said computer to said resource sharing facility;

(f) activating in said resource sharing facility a set of processes in response to at least one request from a respective program running in a respective computer; and

(g) supporting in each process of said set requests from any of said computers

to write a data item in said medium-level cache,

to register interest in a data item,

to write a data item to secondary storage,

to read a data item from said medium-level cache to a given location in the main storage of the requesting computer,

to lock a data item against update or read by other programs,

to unlock a lock,

to perform regular I/O operations, and

to perform any combination of the above requests as a single operation.

13. The method defined in claim 12, further comprising the step of providing each of said computers with a resource sharing facility support facility having a cache buffer control vector in the respective main storage with entries describing states of data items in buffers of the respective high-level cache, each of said requests to register interest in a data item associating a respective one of said entries of a respective cache buffer control vector with a corresponding data item in said medium-level cache. computers and is not shared between said computers.

14. The method defined in claim 13, further comprising the step of periodically transmitting a heartbeat message from each of said resource sharing facility support facilities to said resource sharing facility to signal that a respective computer remains active and connected to the resource sharing facility by a respective intersystem link.

15. The method defined in claim 12, further comprising the step of, upon each writing of any data item from one of said computers into said medium-level cache by a write request, invalidating or refreshing corresponding copies of the same data item in all of said high-level caches of all of said computers that have registered interest in that data item except for the computer which has issued the write request.

16. The method defined in claim 12, further comprising the step of, upon each writing of any data item from one of said computers into said secondary storage by issuance of a write request, optionally writing a copy of the data item in said medium-level cache and invalidating and refreshing copies of said data item in all of said high-level caches of all of said computers that have registered interest in that data item except for the computer which has issued the write request.

17. The method defined in claim 12 wherein the shared data items are selected from:

structured data of a database management system data set,

structured data of a nondatabase management system data set, and

unstructured data items.

18. The method defined in claim 17 wherein said structured data is selected from the group which consists of:

data of an MVS DB2 database,

data of an IMS database,

data of an ADABAS database,

date of a VSAM file,

data of a DB2/6000 database,

data of an Oracle database,

data of a Sybase database,

data of an Informix database,

data of an Ingres database,

data of a Microsoft®

SQL server database,

data of an SQL/DS database,

data of a CA-IDMS database,

data of a DataCom/DB database,

data of IAM files, and

data managed by any other DBMS.

19. The method defined in claim 12 wherein said resource sharing facility database is implemented by a management system directly in the resource sharing facility and is operated for maintenance and use of queues needed for online transaction processing and batch job execution.

20. The method defined in claim 12, further comprising the step of supporting shared database journal files in the resource sharing facility.

21. The method defined in claim 12, further comprising the step of effecting transaction and job load balancing operations using the resource sharing facility for sharing computer independent session context data and job context data.

22. The method defined in claim 12, further comprising the step of implementing a management system for resource sharing facility database in at least one of said computers.

23. The method defined in claim 22 wherein data maintained by said resource sharing facility database management system is stored in said main storage of said one of said computers and is not shared between said computers.

24. The method defined in claim 12, further comprising the steps of:

generating a cross-notification message selectively invalidating a respective high-level cache, refreshing a respective high-level cache and announcing a change in status of a data item associated with the respective high-level cache, by said resource sharing facility upon a change to any data item in said medium-level cache; and

sending said cross-notification message to at least some of said computers that have registered an interest in the respective data item.

25. The method defined in claim 24, further comprising the step of:

posting with the cross-notification message an event control block that has previously been created by the interest registration, in the computer that receives the message.

26. The method defined in claim 12 wherein said computers are selected from the group which consists of IBM®

mainframes and compatible, UNIX machines, NT machines, PCs and compatibles or any other type of computer and combinations thereof.

27. The method defined in claim 12, further comprising the steps of:

reading data items from said medium-level cache with a read request; and

locking data items read from said medium-level cache against concurrent update or read in the medium-level cache, with the same read request.

28. The method defined in claim 12 wherein at least some of the computers are heterogeneous and are connected to a heterogeneous computer system secondary storage controller and wherein some shared data resides in said heterogeneous computer system secondary controller which is used to read and write but not lock said some shared data and where mapping between the portion of said some shared data which is kept also in the medium level cache and its permanent copy in secondary storage of the heterogeneous computer system secondary controller is not delegated to said resource sharing facility.

29. A computer system capable of sharing data items and locks and of supporting remote procedure calls and coordinated process groups, comprising:

a plurality of computers each having one or more central processing units, respective main storage associated with the central processing unit or units of each computer and respective high-level caches maintained in the main storage of the respective computer;

a resource sharing facility with a medium level cache and a resource sharing facility database and capable of selectively storing a set of data items in said medium level cache and in said resource sharing facility database; and

high-speed intersystem links connecting said resource sharing facility to all of said computers so that said resource sharing facility is accessible to all of said computers, said resource sharing facility being provided as a controller for at least one secondary storage, said resource sharing facility being provided with control information and logic needed for mapping medium level cache data to said secondary storage and secondary storage data to said medium level cache for each data item in said medium level cache, cache data maintained in the resource sharing facility and persistent data residing in said secondary storage being synchronized by delegation of input/output operations from the computer to the resource sharing facility.

30. The computer system defined in claim 29, further comprising zero or more coupling facilities with medium level cache that is connected to at least two of said computers in addition to one or more of said resource sharing facilities.

31. The computer system defined in claim 29 wherein at least some of said computers are heterogeneous and are connected to a heterogeneous computer system secondary storage controller and wherein some shared data resides in said heterogeneous computer system secondary storage controller which is used to read and write but not lock said some shared data and where mapping between the portion of said some shared data which is kept also in the medium level cache and its permanent copy in secondary storage of the heterogeneous computer system secondary storage controller is not delegated to said resource sharing facility.