Versioning of message formats in a 24-hour operating environment

• US 5,027,316 A
• Filed: 08/08/1989
• Issued: 06/25/1991
• Est. Priority Date: 09/11/1986
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Frantz, Billy J., Hellstrom, Sven E.
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): Chun, Debra A.
• Application Number:
  US07/391,093
• Time in Patent Office: 686 Days
• Field of Search: 364/200 MS File, 364/900 MS File
• US Class Current: 710/11
• CPC Class Codes: G06F 13/387 for adaptation of different data processing systems to different peripheral devices, e.g. protocol converters for incompatible systems, open system
• CPC Class Codes: G06F 16/1767 Concurrency control, e.g. optimistic or pessimistic approaches
• CPC Class Codes: G06F 16/2329 using versioning
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores

Abstract

Method and apparatus for maintaining currency in interrelated data which is referred to but not varied by users in a multi-user system which is on-line for 24 hours a day. Reference is had to the data through a data indexing construct available to each user. Currency is maintained by adding updated elements of the data when required by changed external conditions and storing the updated elements in the multi-user system concurrently with the previous version. A new indexing construct is built for the updated data version and users are switched to the construct when access to prior data elements is no longer required. Since only a single variable relating to the indexing construct must be changed in order to switch a user, an atomic operation is utilized to perform the switching.

Claims

1. In a continuously-operating, multi-tasking computer system which includes a central facility with a data storage facility, a plurality of user terminals and a message exchange facility for conducting a message exchange between the control facility and the user terminals in which a message exchanged between the control facility and a user terminal has a format established by a message format descriptor, a method for changing message format descriptors without stopping operation of the computer system, comprising the steps of:

defining and storing list of message format descriptors in the data storage facility;

creating an initial index to the list;

for a user terminal, establishing a user terminal block at the central facility, the user terminal block including a single index pointer conditioned to identify the initial index;

during a first message exchange between the central facility and the user terminal which includes transfer of messages formatted according to message format descriptors included in the list accessing the list by way of the initial index in response to the index pointer;

modifying the list of message format descriptors;

creating an updated index to the list;

initiating a second message exchange between the central facility and the user terminal, the message exchange including transfer of messages formatted according to modified message format descriptors included in the list;

in response to initiating the second message exchange, executing an atomic instruction to condition the index pointer to identify the update index; and

during the second message exchange, accessing the list by way of the updated index in response to the index pointer.

2. The method of claim 1, wherein each step of creating includes:

creating a hashing table which includes a plurality of hashing entries, each hashing entry having a predetermined hashing correspondence with one or more of the message format descriptors and each having entry including a header pointer; and

creating an intermediate data structure which consists of a plurality of header entries, one or more of these entries being indicated by a respective header pointer and each header entry including at least a synonym field for storing the identity of another header entry, a name field for storing the identity of one of the message format descriptors, and a list pointer for indicating the storage location in the data storage facility where the message format descriptor identified in the name field is located.

3. A lock-free method for assuring independent access to different versions of a message descriptor in a list of message descriptors maintained in a data storage facility of a central computer facility to which a plurality of user terminals are connected, the central computer facility including a message exchange facility for conducting message exchange between the control computer facility and the user terminals in which message formats are defined by message descriptors, the method comprising the steps of:

defining and storing a list of message descriptors in the data storage facility;

creating a first index to the list of message descriptors;

providing a terminal block for user terminals exchanging messages with the central facility, each terminal block including a single index pointer conditioned to identify the first index;

in response to exchange of a first message between a user terminal and a central facility, accessing the list of message descriptors by way of the first index in response to an index pointer in a terminal block for the user terminal;

updating a message descriptor in the list of message descriptors and adding the updated message descriptor to the list of message descriptors;

creating a second index to the list of message descriptors including the updated message descriptor;

in response to exchange of a second message between the user terminal and the central facility, atomically altering the index pointer in the terminal block to identify the second index; and

accessing the list of message descriptors by way of the second index in response to the pointer.

4. The method of claim 3, wherein each of said creating steps includes:

creating a hashing table which includes a plurality of having entries, each having entry having a predetermined having correspondence with one or more of the message descriptors and each hashing entry including a header pointer; and

creating an intermediate data structure which consists of a plurality of header entries, one or more of these entries being indicated by a respective header pointer and each header entry including at least a synonym field for storing the identity of another header entry, a name field for storing the identity of one of the message descriptors, and a list pointer for indicating the storage location in the data storage facility where the message descriptor identified in the name field is located.