Multi-processing financial transaction processing system
A financial transaction processing system is disclosed, wherein substantial processing efficiencies are provided with, additionally, a substantial decrease in the size of the executable code. Each transaction processed by the transaction processing system is described by a transaction data descriptor that includes a series of subtransaction data descriptions of actions that can be performed independently of one another. Thus, complex transaction processing logic is substantially removed from the executable code, and instead such transaction data descriptors are processed interpretatively. Moreover, the independence of the subtransactions allows the subtransactions of a transaction to be processed in parallel when performed on a multiprocessor computer. Additionally, the transaction processing system provides account balancing enhancements in that there are control columns in various data tables that are automatically updated during transaction processing so that by comparing control column totals, an indication of the integrity of current financial records is provided. Additionally, the transaction processing system provides full auditability in that any changes to financial data can be traced for any effective period of time into the past so that auditors can periodically perform a full audit of the financial transaction data retained by the transaction processing system.
The present invention relates to a financial transaction processing system, and in particular, to such a system that is capable of decomposing transactions into subtransactions and multi-processing subtransactions simultaneously.
BACKGROUND OF THE INVENTIONComputerized data processing systems for processing financial transactions have become increasingly more complex as further strides toward automation have occurred. Such complexity has generated a number of related difficulties for the financial data processing industry. In particular, complex financial transaction processing systems may have subtle programming defects or errors that may go unnoticed for long periods of time before the extent of the problems thereby generated are fully recognized. For example, the number of positions allotted for the dating of transactions has recently been problematic, wherein the dates for the millennium starting at the year 2000 can be problematic for many financial transaction processing systems.
In addition, such complex financial transaction processing systems also are typically incapable of being fully audited. That is, it is common practice in the financial data processing industry to provide only partial auditability in that it is generally believed that the amount of data required to be stored for full auditability is so large as to not be cost effective.
Further, in many circumstances, the rate of transaction increase is becoming problematic in that progressively larger computers are required for processing financial transactions at an acceptable rate. This problem is exacerbated by the fact that such transaction processing systems are not architected for use on multi-processing machines having a plurality of processors. Thus, the advantages of parallel-processing computers cannot be fully utilized by such systems.
Accordingly, it would be advantageous to have a financial transaction processing system that alleviates the above difficulties, and that additionally, provides flexibility to adapt to the changing business needs of business enterprises so that the transactions processed and the respective reports generated may be modified easily according to business constraints and demands.
SUMMARY OF THE INVENTIONThe present invention is a financial transaction processing system that achieves substantial increases in auditability and processing efficiency. In particular, the present invention provides auditable trails or history in a number of different ways. For example, financial data within transactions is used in the present invention to update various control fields in different tables or files so that cross-checks of system financial integrity can be performed for assuring that, for example, cash fields, total units fields, and cost fields balance appropriately across system data tables provided by the present invention. Additionally, the present invention provides a full range of auditable history files for each system data table having information that is required during auditing.
The present invention also performs financial transaction processing using a novel computational paradigm. That is, the financial transaction processing system of the present invention has an architecture wherein financial transactions can be decomposed into corresponding collections of independent subtransactions, such that for each input transaction, the corresponding collection of subtransactions are performed by operations that are independent of one another. Thus, the subtransactions can be performed in any order, including in an overlapping fashion, such as may occur during multiprocessing of these subtransactions on a computer having multiple processors.
Further, note that each of the subtransactions is described by a relatively short (e.g., less than 8 characters) text string that can be straightforwardly interpreted as an operation (e.g., either plus or minus) together with a series of operands, in particular, a first operand having a value to be used in modifying a data table field (column) specified by a second operand. Such high level descriptions of subtransactions provide both compact conceptualization and a reduction in the total size of the executable code for the present invention. Accordingly, when one of the subtransactions is performed, not only is its corresponding operation performed on the operands, but additionally, control fields such as those mentioned above are updated appropriately in various data tables for the present invention to enhance auditability of the financial data resulting from the transaction processing. Further, note that since the subtransactions are independent of one another and their executable code is relatively small, there is no need for lengthy and complex flow of control transaction processing modules. That is, the size of the code for the present invention may be up to 100 times smaller than many prior art transaction processing systems. Accordingly, this has a substantial positive impact on the efficiency of the present invention in that the swapping of program elements in and out of primary computer memory is substantially reduced.
In another aspect of the present invention, the financial transactions of a plurality of business enterprises can be processed in an interleaved manner. In particular, since the present invention is substantially data driven, including the descriptions of the transactions and their related subtransactions, the present invention can be easily modified to incorporate both different or updated versions of transactions and associated data tables for an existing business enterprise (e.g., also denoted “licensee” hereinafter). Additionally, the transactions and related data tables for an entirely new or different business enterprise (licensee) may be straightforwardly incorporated into the present invention so that its transactions can be interleaved with the transactions of other business enterprises. Thus, transaction processing may be performed by the present invention for business enterprises having different transactions, different account record structures and differently organized general ledgers substantially without modifying the program elements of the transaction processing system.
For example, the present invention can be used to simultaneously process transactions for:
-
- (1) a single software application such as an investment management or telecommunications billing system,
- (2) multiple disparate software applications such as investment management, and telecommunications billing, paying agencies, etc., all with disparate definitions.
Accordingly, the present invention may be viewed as a software engine, or a user-definable transaction processing tool that can be adapted to a variety of industry specific software application needs without changing the actual program code. That is, by surrounding the present invention with application specific software for inputting transaction data to the multi-processing financial transaction processor of the present invention and retrieving data from the multi-processing financial transaction processor of the present invention, a particular business enterprise can have substantially all of its financial records in condition for auditing on a daily or weekly basis.
The present invention may be further characterized along the following dimensions: flexibility, auditability, multiprocessing, efficiency and size, these dimensions being discussed, in turn, hereinbelow.
Flexibility is achieved by permitting a business enterprise to define:
-
- (1) a series of “reference” tables (also denoted “master tables”) that describe the appropriate management decision-making, accounting structure, and regulatory information for the specific application;
- (2) a series of audit controls and system procedures that provide for complete control of all processing and prevent the overwriting of any original data;
- (3) a series of institutional and customer reporting files, known as the “driven” tables; and
- (4) the specific processing content of each individual transaction to be processed via a series of table definitions, known as the “driving” tables.
Thus, transactions may be customized according to the business needs of a business enterprise.
Auditability is achieved by:
-
- (1) providing separate control columns for cash, units and cost basis (if any) in detail records generated and stored for each financial transaction;
- (2) repeating these three control columns, or variations thereof, in at least three different tables so that subsequent summations of each of the four tables will result in similar balances and thus prove that no critical data has been lost in the course of processing, as one familiar with auditing and financial transactions systems will understand;
- (3) adding appropriate data columns:
- (a) to each reference table or master row for maintaining a history of the effects of add, change and delete commands in a current database as well as an archive database;
- (b) to each original file record (i.e. table row) that represents an add to a current database as well as the periodic archive and purge to a permanent database;
- (c) to tables for retaining transaction processing data representing error identification, error negation and error correction.
Thus, auditabilty of transaction records is achieved by four sets of files for a specific period. These are: (a) a snapshot of all the reference files at the end of the period; (b) snapshots of a history file for each master table, wherein the corresponding history file (table) contains all changes to the master table during the specific period; (c) a snapshot of all financial transactions for the specific period, and (d) a snapshot of all of the “driven” tables at the end of the period.
Multiprocessing is achieved by:
-
- (1) decomposing the processing of the present invention into a series of separate and independent subprocesses that may be simultaneously performed on any number of simultaneous processors, and
- (2) decomposing input transactions into a series of subtransactions that are processed by independent processes, which may be executed in any particular order, with complete auditability.
For example, multiprocessing can be achieved by allocating the next prescribed subtransaction process to the next available processor.
Efficiency is achieved by:
-
- (1) Defining and utilizing only four standard processing models that perform all prescribed functionality and auditability of the present invention. The models are:
- (a) Processing Model 1 provides an architecture for maintaining historical transaction data so that financial changes can be traced through time;
- (b) Processing Model 2 provides an architecture for automatically maintaining data columns such as Units, Debits and Credits for cross checking table sums to assure that the financial records for a business enterprise balance;
- (c) Processing Model 3 provides an architecture for automatically maintaining financial records relating to financial instruments such as stocks, bonds, real estate, etc.; and
- (d) Processing Model 4 provides an architecture for producing a common processing format for maintaining customer and institutional data tables.
- (2) Defining only four primary program modules for controlling functionality of the present invention, these modules being:
- (a) a transaction processing controller module for receiving transactions to be processed, and controlling the processing thereof;
- (b) a preprocessor and decomposer module for determining the validity of a received transaction, assuring that all data tables and rows thereof are available for processing the transaction, and retrieving the appropriate subtransactions data descriptions to be processed;
- (c) a subtransaction scheduling module for scheduling instantiations of the subtransaction processing module on each of one or more processors; and
- (d) a subtransaction processing module for performing each subtransaction retrieved by the preprocessor and decomposer module.
- (3) Utilizing a number of software switches to control which tables within collection of “driven” tables are to be updated when a specific type of transaction is to be processed.
- (1) Defining and utilizing only four standard processing models that perform all prescribed functionality and auditability of the present invention. The models are:
Thus, by providing a small number of processing models, decomposing input transactions, and supplying only the necessary subtransaction descriptions, the reliability of the transaction processing system of the present invention is substantially increased.
The software for the present invention is small in size (both source code and object code) due to the following:
-
- (1) defining business enterprise financial data processing methods, accounting structures, and regulatory definitions as data rather than program code;
- (2) reducing the processing content to a series of individual transactions; and
- (3) reducing all financial transactions to a collection of subtransactions wherein each subtransaction includes an operator and two or more operands in an 8-character string.
Thus, the financial processing by the present invention may be performed on several transactions at a time, one transaction at a time, or different processors within a multiprocessor context. Or, the subtransactions for a specific transaction may be spread over several simultaneous processors. This means that the business enterprise is afforded a large number of options in tailoring the present invention.
Hence, by defining the accounting structure and processing functionality as data rather than actual program code, the size of the total code required to process a specific industry application may be substantially reduced compared to prior art transaction processing systems. For example, the executable code for the present invention may be less than one megabyte (1MB). Thus, since the secondary cache attached to each processor in multiprocessing personal computer servers can be one megabyte, substantially the entire executable for the present invention can be provided to each processor. Thus, the positive impact on total system efficiency is believed to be substantial in that secondary cache is typically about four times faster than normal cache, so productivity gains of about three-hundred percent would not be unreasonable. In other words, the executable code for the present invention can reside in the secondary cache of each processor, thereby allowing the off-loading of any processing function to any processor with relative ease. Additionally, given that a typical RAM memory for a personal computing devices is 16 megabytes, it is believed that such a device will have the capability to process the back office financial transactions of a major money center financial institution or communications billing system.
Additional features and benefits of the invention will become evident from the detailed description and the accompanying drawings contained herein.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGS. 4-A through 4-E illustrate the steps of a flowchart for initializing the database tables of the present invention for a new business enterprise licensee that is to have its financial transactions subsequently processed by the present invention.
FIGS. 7-A through 7-D show the high level steps performed by an embodiment of the preprocessor and decomposer 54 of
FIGS. 8-A and 8-B show the steps of a flowchart for obtaining indexes or pointers to particular rows of a general ledger table wherein the rows are used in processing a transaction.
FIGS. 9-A and 9-B show the steps for a flowchart of an embodiment of the subtransaction processing module 64 (
(a) transaction processing controller 52 for: (i) receiving transactions 58 from business enterprises, (ii) controlling the processing of such transactions, including the scheduling of subtransactions to be performed, and (iii) writing of transaction details to, for example, a transaction journal file or table;
(b) a transaction preprocessor and decomposer 54 for initially receiving a transaction 58 from any one of a plurality of business enterprises as shown, wherein the preprocessor and decomposer 54 decomposes transactions into subtransactions;
(c) a subtransaction processing module 64 for performing the instructions for each subtransaction determined by the transaction preprocessor and decomposer 54. In particular, the subtransaction processing module 64 utilizes a collection of subtransaction programmatic data descriptions 66 that can be independently scheduled and performed for processing each transaction 58 provided to the transaction processing system 50;
(d) a subtransaction scheduler 62 for scheduling the execution of each subtransaction output by the preprocessor and decomposer 54;
(e) a collection of databases 70 containing financial information for each of the one or more business enterprises. Note that the term “database” in the present context includes both the data therein as well as database management functional elements and data structure definitions.
Another illustration of the present invention is provided in
Referring still to
-
- (a) determines, using input from the business enterprise databases 70, whether all necessary data for performing the transaction is available and otherwise rejects the transaction without performing any portion thereof. In particular, the transaction preprocessor and decomposer 54 determines that all data tables to be accessed are available;
- (b) retrieves the data needed to perform the transaction;
- (c) checks to determine that the transaction operation(s) requested is available, and that the transaction is legitimate to be performed on the data for the input transaction 58;
- (d) retrieves the subtransaction data descriptors for decomposing the input transaction 58 into subtransactions.
Accordingly, the preprocessor and decomposer 54 retrieves into the working storage 72 (shown inFIG. 3 ) of a host computer (not shown), upon which the transaction processing system 50 is operating, substantially all data and table rows that are necessary to process the transaction 58. Additionally, note that as one skilled in the art will understand, if some portion of the required data to process the transaction is unavailable, then the preprocessor and decomposer 54 terminates processing and subsequently writes appropriate error messages and/or details of the transaction into the reject table 74 (FIG. 2 ).
Assuming that the preprocessor and decomposer 54 successfully performs the gathering of information for the decomposing of the transaction into subtransactions appropriately, then control is returned to the transaction processing controller 52, wherein this controller then writes the details of the transaction to the transaction journal 78 along with identification data uniquely identifying the transaction (e.g., a transaction sequence number and/or time and date stamp). Following this, the transaction processing controller 52 invokes the subtransaction scheduler 62 for scheduling the performance of each subtransaction by an invocation of the subtransaction processing module 64. Note that it is an important aspect of the present invention that since the subtransactions can be processed independently of one another for a given transaction, instantiations of the subtransaction processing module 64 can be executed in substantially any desired order. In particular, such instantiations of the subtransaction processing module 64 can be performed concurrently, thus providing a substantial increase in transaction processing efficiency when such concurrency is provided on a computer having a plurality of processors.
Given that a subtransaction is performed successfully by the subtransaction processing module 64, various accounting tables within the transaction processing system 50 are updated. In general, each subtransaction conceptually indicates a single operation of either plus or minus that is to be performed with two operands also indicated in the subtransaction. That is, the first operand indicates the data to be added or subtracted from a particular field or column of a table row identified by the second operand. Additionally, each subtransaction updates other tables within the transaction processing system 50 automatically in order to provide consistency among the data tables so that: (a) substantially on-line account balancing capabilities can be performed, and (b) full auditability of the records of the business enterprise providing the transaction can be facilitated by retaining history records of table updates, as will be discussed with reference to “master table transaction cluster processing” described hereinbelow. Accordingly, each subtransaction processed by an instantiation of the subtransaction processing module 64 may update a plurality of the data tables contained in the collectively labeled database 70. Note that for one skilled in the art of transaction data processing and accounting, the names provided to the tables are indicative of their information content and structure. However, for clarity, substantially all of the tables for the present invention will be discussed in detail and/or illustrated hereinbelow.
The subtransaction processing module 64 processes subtransactions derived from three general categories of transactions that may be input to the present invention. That is, there may be input transactions for each of the following types of financial transactions (1.1) through (1.3) hereinbelow.
(1.1) Transactions related to exchanges of funds such as cash debits and credits for accounts of a particular business enterprise are provided. At a high level, the tables related to this functionality include the account master table 84 (
(1.2) Transactions related to additional or customized accounting for clients having accounts in the account master table 84 are provided. For example, in addition to providing the functionality of the transactions described in (1.1) immediately above, a customer income statement (income/expense) table 96 may be provided with client account and transaction information related to income and expenses for tax purposes. Additionally, a customer cash flow (receipts/disbursements) table 100 is also provided for recording any account transaction information related to receipts and disbursements in client accounts. Further, a customer performance measurement table 104 is also provided for retaining client account performance information related to the performance of client portfolios in comparison to investment indexes such as the Dow Jones Industrial Average, the S&P 500, etc. Note that these tables will be discussed and/or illustrated hereinbelow.
(1.3) When transactions are additionally related to financial instruments other than cash, debits and credits, such as portfolio management wherein there is buying and selling of equities, income derived from equities, and trade settlements related thereto. Further, note that these additional capabilities also provide the same degree of flexibility, adaptability and simplicity as provided in relation to the transaction processing capabilities discussed in (1.1) and (1.2) immediately above. That is, financial equity transactions of various types and for various business enterprises may be easily modified and/or added or removed from the transaction processing system 50 of the present invention, since these transactions are also described by transaction data descriptors consisting of a collection of subtransactions that are capable of being performed in substantially any order that is determined by the subtransaction scheduler 62.
Accordingly, in providing the functionality for the transactions related to portfolio management, the preprocessor and decomposer 54, upon being invoked by the transaction processing controller 52, also retrieves into working storage (as shown in
(2.1) Original add module 114 for processing a subtransaction related to the addition of further financial instruments to a portfolio such as occurs when securities are bought and must be added to a given account.
(2.2) A reverse of add module 118 for reversing an addition of financial enterprises to a particular account portfolio. Note that this module is typically activated when financial enterprises are inadvertently added to an incorrect portfolio account.
(2.3) An original sell module 122 for processing subtransactions related to selling financial enterprises within a given account portfolio.
(2.4) A reversal of original sell module 126 for reversing the affects of an inadvertent sell of financial enterprises within an account portfolio.
These four modules 114- 26 update the tables labeled collectively as 70B. In particular, the processing performed herein and the tables updated herein are described below.
Major Programs and FunctionalityMajor Programs
The N_gine transaction processing system contains four major programs. These are:
-
- (1) Transaction Processing controller 52
- (2) Transaction Preprocessor and Decomposer 54
- (3) Subtransaction Processing module 64
- (4) Subtransaction Scheduler 62
Program Functionality
The purpose of the Transaction Processing controller 52
-
- (a) test for incoming transactions and once detected
- (b) execute the Transaction Preprocessor and Decomposer 54 and then
- (c) execute the Subtransaction Processing module 64 for each transaction.
The purpose of the Transaction Preprocessor and Decomposer 54 is to verify
-
- (a) that all information in the transaction is accurate
- (b) that all files and controls are available to properly process the transaction
- (c) that the specific subtransaction processing instructions are loaded into working storage.
The purpose of the Subtransaction Processing module 64 is to
-
- (a) execute all of the subtransactions that have been previously defined for a transaction
- (b) create auditability for every transaction.
The purpose of the Subtransaction Scheduler 62 is to
-
- (a) allocate a specific task to a specific processor
- (b) return processing to the Transaction Processing controller 52.
The present invention may be described as “Table-Driven Transaction Processing”. That is, the present invention permits the processing of virtually any type of user-definable transaction by defining the processing for such transactions as data descriptors that are interpreted in real time and dynamically as needed for processing corresponding transactions. Accordingly, the transaction data descriptors are denoted as “driving data” and are defined by the transaction processing master table and the transaction master table. That is, the transaction master table provides a first initial collection of data for identifying each transaction and the transaction processing table provides the remainder of the data including the subtransaction decompositions. Accordingly, each transaction processed updates an appropriate set of user-definable tables (known as the “driven” data) for completing the processing of the transaction. Since both the “driving” and the “driven” information is expressed as data rather that actual code, the entire functionality of the system can be changed in a straightforward manner.
In the description hereinbelow, the functional components of the present invention are also identified by other naming conventions from the description above. Accordingly, the following table shows the pairing of the functional component identifications above with those also used below:
A. The Magic Number in Software Design is 1. That is,
-
- store data once,
- program data once,
- process data once.
B. Design a total system with the fewest number of processing models. For example,
-
- One model for processing all adds (inserts), changes (updates), and deletes (deletes) for all Master (or Reference) Files (or tables).
- Namely,
- Begin Time
- Number of Transactions
- Number of Acceptances
- Number of Rejects
- End Time.
These variables represent the only true means of measuring actual productivity.
F. For reasons of auditability, never overwrite any original information. Move all original information from data entry (cradle) to data warehouse (grave) without any changes.
G. For reasons of reliability and profitability, system designs should focus on a “large number of small programs” rather than a “small number of large programs”. The result is not only ease of maintenance but also the ability to spread the small programs across a number of simultaneous processors.
H. For reasons of manageability, all system designs should embrace one integrated enterprise-wide standard naming convention for all files (tables), records (rows), and fields (columns).
I. For reasons of portability, use the fewest number of language commands to code the system. Avoid vendor and/or language extensions.
J. For reasons of flexibility, never hard code what can be table-driven.
N_gine Design ConceptsA. Only 4 Processing Models for Financial Services and Telecommunications Applications
-
- 1. Schema
- 2. Units, Debit/Credit
- 3. Assets/Liabilities
- 4. File Maintenance Routine
B. Table-Driven Transaction Processing for maximum flexibility
-
- 1. Number of Transactions
- 2. Name of Each Transaction and Unique Details
- 3. Processing Algorithms (at least 1, up to 20 depending upon complexity)
- 4. Each algorithm has 3 components
- a. Plus (P) or Minus (M)
- b. Operand 1
- c. Operand 2
C. 100% Auditability For Every Transaction by creating
-
- 1. a Detail Record containing all relevant data and
- 2. hash totals of three relevant fields in at least 3 other tables.
D. The 3 relevant fields for calculating all hash totals are:
-
- 1. Cash
- 2. Units
- 3. Cost Basis
E. Basic Relational Database Management System Processing Concepts
-
- 1. Commit/Rollback
- 2. Row Level Locking
- 3. Indexing, ROWID
- 4. Stored Procedures
- 5. Shared Memory
F. Some Financial Services Accounting Systems are not permitted to commingle funds. That is, separate accounting for both income and principal must be provided. Therefore, each account master must have a designated “income posting code” to define the proper processing. Such a code might be: (I) Income Only, (P) Principal Only, (B) Both Income and Principal.
N_gine's Basic TablesLicensee Profile (The Licensee “Reference” or “Master” Tables)
- LM The License Master table contains the necessary information to process any type of licensee using either single or multiprocessing computers.
- LU The Licensee User Master identifies different users for the disparate systems that may be processed simultaneously.
- LT The Licensee Account Type table contains the necessary information to process any type of account be it for a pension trust account, a communications account, or a corporate subsidiary.
- LD The Licensee Default Definition table the default definitions for cash, units, and cost basis controls for total system control.
- LL The Licensee General Ledger Definition is a list of all of the acceptable entries for the General Ledger. That is, it provides a framework for processing any type of accounting controls for any set of account types.
- LS The Licensee Diversification Scheme contains a three level classification scheme for reporting an decision-making purposes for any set of assets and liabilities.
- LP The Performance Measurement Group Master contains a three level classification scheme for measuring the performance of different investment groups.
- LN The Licensee Summary Name Master contains a list of the entries on any type of Income Statement and Cash Flow Statement.
- LW The Licensee Wholesaler Master contains name, address, sales volumes, etc. wholesalers of communications services.
- LR The Licensee Reseller Master contains name, address, sales volumes, etc. for resellers of communications services.
Account Profile (The Customer “Reference” Tables) - AO The Account Objectives Table contains the different types of account objectives, such as income, growth, capital preservation, etc.
- AL The Account Jurisdiction contains the different types of legal relationships, such as broker, agent, trustee, advisor, etc.
- AJ The Account Jurisdiction contains the different types of legal jurisdiction, such as federal law, state law, foreign law, etc.
- AR The Account Representatives Table houses the different representatives, their names and communication addresses.
- AN The Account Registration Names is a list of legal names used in security settlement.
- AM The Account Master table provides all of the necessary information to process any type of account by linking the Account Objective, Account Jurisdiction, Legal Capacity, Profit Center, Account Representative, and Registration tables plus other relevant data for reporting content and reporting cycles.
- AC The Account Communications Links links the Account Number for Financial Services to the account numbers for communications services so that all information can be contained in one reporting scheme.
- TM The Transaction Master table provides all of the information to process any type of transaction, excepting the specific processing algorithms.
- TP The Transaction Processing table provides all of the specific processing algorithms for any type of transaction master. The Transaction Master and Transaction Processing tables provide all of the necessary information to process any type of transaction.
- TR The Transactions—Recurring Table (TR) contains the necessary information for automatically processing any type of transaction on a recurring basis.
- EM The Entity Master table provides all of the necessary information to process any type of financial entity.
- EA The Entity Attribute table joins all relevant diversification (known as type, group, and class), general ledger (known as accounting control numbers), and performance group (known as type, group, and class) data into one table for only one access seek.
- ET The Entity Transaction table links specific transactions to specific entities, such as BG (Buy Government) for a US Treasury Note, BF (Buy Tax-Free) for a tax-free bond, BE (Buy Equity) for common stocks, etc. Note: It is the correct assignment of such transactions to such entities that permits the proper accumulation of data for income tax purposes.
- SG The System General Ledger contains all of the information to process any type of institutional accounting control.
- SJ The System Transaction Journal Table contains all of the transactions and all of the details for each transaction for a specific accounting period.
- ST The System Trade Settlement Table contains all of the automatically generated offset transactions for Buys and Sells
- SS The System Summary Table contains a record for each execution of the system with the Begin Time, End Time, Number of Total Records Read, Number of Accepts, Number of Rejects, etc.
- SR The System Reject Table contains a list of all transactions rejected for whatever reason.
- SC The System Transaction Count Table contains the number of each type of transaction processed on any given transaction.
- CS The Customer Income Statement contains all revenues, expenses, and profits or losses for all customer accounts.
- CF The Customer Cash Flow Statement contains all receipts and disbursements for all customer accounts.
- CB The Customer Balance Sheet table contains all assets and liabilities for all customer accounts.
- CG The Customer Capital Gains table contains all of the realized capital gain details for all customer accounts.
- CI The Pending Income table contains all of the pending income, such as interest or dividends, for all accounts.
- CA The Pending Capital Adjustments table contains all of the pending capital adjustments, such as stock splits, stock dividends, mergers, acquisitions, etc., for all accounts.
- CP The Performance Measurement contains all of the periodic performance records for all customer accounts.
Since every transaction is recorded in a detail record plus hashed to three other control tables, the control values of cash, units, and cost basis are added to like values in the following control tables:
Account Master, System General Ledger, and Entity
Attribute tables.
For other reports such as the Income Statement and the Cash Flow Statements, the Performance Measurement table is used as a control table instead of the General Ledger.
The present invention includes four computational processing models (process models 1 through 4) for processing financial transactions and assuring full auditability and traceability.
The purpose of Process Model 1 (
A current database 300 (
Eight tables (i.e. tables 312, 316, 320, 324, 328, 332, 336 and 340, of
Number of Modifications
12 Control Fields per master table
A sequence number generator
A process flow methodology for add, change, and delete of data table rows.
The operation of Process Model 1 is as follows:
1) Normal Updating to current database 300
- 2) Periodic updating to the archive database 304 at the end of a pre-determined time period. That is,
- (a) archive snapshots of the archive master 312 in the current database 300 to the master in archive database 304;
- (b) archive the archive history 332 in the current database 300 to the master history 340 in the archive database 304;
- (c) purge the history table 332 in the current database 304.
The purpose of Process Model 2 (
The methodology of process model 2 utilizes a data-driven transaction processing strategy, wherein the manner in which a transaction is processed is determined by retrieving appropriate control data for processing a given input transaction. Thus, the present model provides the ability: (a) to process like systems (such as financial services systems) with different transaction definitions and accounting requirements (such as commercial banking, broker/dealers, mutual funds, insurance systems) and different debits and credits and/or (b) unlike systems (such as telecommunications systems) with disparate definitions (such as landline, wireless, satellite, cable systems) within the present invention at the same time.
The purpose of Process Model 3 (
-
- “Detail Record Maintenance”, that is, the ability to process transactions for similar business enterprises (such as portfolio management systems) relating to various financial instruments (such as disparate assets and liabilities) and/or transactions for dissimilar business enterprises (such as portfolio management systems, paying agencies, stock transfer systems) with disparate languages (such as English, Spanish, French, or German) and disparate definitions (such as management philosophy, accounting, and operating nomenclature) and unlike financial instruments (such as assets and liabilities) within the same software at the same time.
- The ability to decompose, allocate, process, and audit each financial instrument transactions with 100% auditability.
- The current databases 300 (for additions, negations and corrections) and the archive databases 304(read only);
- Sixteen data tables (some of which are shown in
FIGS. 2A-2B ) plus a sequence generator; - 12 control fields appended to the master tables for tracing master table changes;
- One transaction three hash totals (mostly using AM, EA, and PM tables);
- 4 currency fields;
- Sequence number generation;
- Reversing/reversed by detail;
- Processing flow for additions, negations, and corrections.
The purpose of Process Model 4 is to create a single methodology for performing file maintenance including: creating a record (row) containing the initial data in a file (table) or modifying the initial data within an existing record (row) within a file (table) or deleting a current record (row) from a file (table) in any software application on computing configurations using simultaneous processors. Where the term, “Details”, hereinbelow represents the identity of the specific financial transaction, the methodology of the process model 4 is provided by programs such as the following:
Accordingly, the methodology of process model 4 defines:
-
- (a) A current database (for additions, negations and corrections) and archive database (Read Only)
- (b) ADD or SUBTRACT;
- (c) Initial tests for values;
- (d) Special deletion criteria;
(e) Tests for action;
-
-
- INSERT or UPDATE;
- DELETE or UPDATE;
- INSERT INTO Reject Tables;
Processing Model 1:
-
Processing model 1 is a method for processing changes to files (or tables) denoted as master or reference tables (files) wherein these tables retain fundamental information that is not derivable from other tables. In particular, processing model 1 processes changes to master tables in an automated manner without losing historical financial information. Accordingly, 100% auditability of all data changes is able to be achieved.
The method of achieving this goal uses an architecture denoted as “Master Transaction Cluster Processing” (MTCP). MTCP is based on the premise of creating a logical flow of all original information from data capture (data entry) to permanent data repository (data warehouse) by replacing single master files (or tables) with a cluster of files (or tables). Therefore, MTCP addresses the complete life cycle of all information relevant to organizational decision-making. MTCP is targeted for use in the automatic generation of program code for multiple large-scale real-time transaction processing applications (such as securities trading, telecommunications billing, and work management) on multi-processing computers (using 4, 8, 16, 32 processors), where control is not only an increasing complex issue but an absolute necessity for future competition.
The circumstances leading to the invention of Master Transaction Cluster Processing are:
-
- a) Prior art financial transaction software architecture lacks the ability to identify transactions by table, transaction date, transaction number, and the person authorizing the transaction.
- b) Prior art financial transaction systems typically use only one table to contain all Master Information (i.e., non-derivable information) and the data in this table is overwritten, thereby losing historical information. Cases in point would be a record of all of the past mailing addresses or processing instructions for a specific customer.
- c) Without 100% retention of an organization's vital information, management has no idea of the accuracy of the information being used for decision-making purposes.
- d) The Year 2000 problem, know as Y2K, is proving that past software applications designs have reached technological limits and current maintenance costs are inordinately expensive.
- e) Competitive pressures are mounting for higher quality software with lower software development and maintenance costs. Totally new architectures for applications software is in great demand.
- f) The ComputerWorld article, “Information: America's Favorite Investment,” by Paul Strassman, ComputerWorld Magazine, Aug. 5, 1996, states that over 1100 companies are spending more on automation annually than the net worths of their respective companies.
- g) The Standish Report as described in Development Patterns, InfoWorld Magazine, Feb. 3, 1997, p. 56, states that the success rate of Business Process Reengineering has increased from 16% in 1994 to only 27% in 1996.
Note, in the book “Oracle Design”, Ensor & Stevenson, O'Reilly Press, it is a recommended practice to compromise data retention rather than achieve 100% auditability. Today's hardware costs suggest otherwise.
The advantages of the present invention over the approaches discussed above are:
-
- to provide 100% auditability which offers business management the capability to exercise its fiduciary responsibility to its stockholders and Board of Directors,
- to capture, maintain, and ensure the integrity of all vital information for business enterprise decision-making purposes, and
- to preserve such information consistent with business enterprise-defined data retention cycles. Additionally, the present invention allows accountants to certify in business enterprise annual reports that all vital corporate data is being properly preserved.
A detailed description of Master Transaction Cluster Processing corresponding to model 1 (the first computational model of the present invention) is as follows.
MTCP Overview
Master Transaction Clustering, or MTCP, performs the following tasks:
-
- a) assigns a unique identifier based on (i) master table identification, (ii) transaction date, (iii) transaction number, and (iv) authorized user, to each transaction that causes a change in the state of a particular record of a master table. That is, if one or more data elements in the record change, then the previous record is written to history, and a new status is assigned to an identifier field used for tracking such changes;
- b) creates a logical flow of data as it is originally entered from its inception (data entry) to its repository (data warehouse). The unique architecture of MTCP replaces the Master File (or Table) within prior art systems with a cluster of Master Files (or Tables), known as a “Master Transaction Cluster”. This cluster is suitable for multiprocessing (or the use of simultaneous processors within a single computer to complete a common job). Hence, MTCP addresses 100% auditability via maintaining the total life cycle of information. Aged information may be deleted from the appropriate tables consistent with user-defined data retention policies;
- c) offers a standard for processing all Master Tables within a total application;
- d) provides a test bed for separately testing each Master Table Cluster under development and all Master Table Clusters in concert;
- e) permits management to report that it is successfully capturing, maintaining, and preserving all critical information for decision-making purposes.
MTCP Scope
Master Transaction Cluster Processing utilizes the following (
-
- a) two databases (i.e., the current data base 300 and the archive data base 304),
- b) sequencing generator 308 having: (i) two external sequence generators; (ii) two internal counters,
- c) eight tables (denoted master table 312, input table 316, summary table 320, reject table 324, accept table 328, history table 332, master archive table 336 and master history table 340), and
- d) twelve additional fields for every row in the master table 312.
MTCP Independence
Master Transaction Cluster Processing of Model 1 is independent of any:
-
- a) application—such as accounts receivable, customer billing, etc.
- b) industry—such as financial services, telecommunication, or work management,
- c) hardware manufacturer—such as Compaq, Digital, HP, IBM, NCR, Unisys,
- d) operating system—such as MS-DOS, UNIX, OpenVMS, MVS, etc.
- e) network—such as Novell, Ethernet, etc.
- f) relational database management system—such as Oracle, Sybase, Microsoft SQL Server, Informix, etc., and
- g) computer language—such as SQL, COBOL, FORTRAN, PL/1, Java, etc.
MTCP Architecture
The Master Transaction Cluster Processing (MTCP) architecture can be used for any application in any industry using any computer language. Within the typical structured processing scheme of input and process, the Master Transaction Cluster Processing focuses solely on the process function. Thus, the method permits users to define input screens and defined output reports.
MTCP Databases
Unlike prior art software system which contain only one table for each set of primary records, Master Transaction Cluster Processing uses eight related tables, or a cluster of tables, to track all information on a cradle to grave basis. The cradle being its point in inception (or data entry), and the grave being its permanent repository (or data warehouse). Consequently, the “Master Transaction Cluster” spans two different databases: one denoted the Current database 300 containing all relevant data for the current processing period and a second denoted the Archive database 304 containing all relevant data for all previous processing periods. The Current database 300 represents the area of high inquiry, and the Archive database 304 represents the area of low inquiry. Consequently, the Current database 300 is normally placed on high-speed internal disk drive and the Archive database 304 is normally placed on less expensive lower-speed CD-ROMs. Note that trailing information in the Archive database 304 may be destroyed consistent with defined data retention policies, statute of limitations, etc.
MTCP Tables
The six tables in the Current database 300 are the
-
- a.) Master Table 312(M) that will contain all records to be maintained.
- b.) Input Table 316 (I) that will contain all records prior to updating.
- c.) Reject Table 324 (R) that will contain all records rejected during processing.
- d.) Accept Table 328 (A) that will contain all records accepted during processing.
- e.) History Table 332 (H) that contain a complete snapshot of all records prior to updating.
- f.) Summary Table 320 (S) that contains the results of a specific processing operation.
and the two tables in the Archive database 304 are the: - g.) Master Archive Table 336 that contains snapshots of the master table 312 at the end of each processing period.
- h.) Master History Table 340 that contains a history of the master table 312 changes during a current processing period.
Note that the Master Table (M), Input Table (I), Reject Table (R), the Accept Table (A), the History Table (H) in the same “Master Transaction Cluster” share the same number and order of data elements consisting of alphabetic, numeric, and date items. Alternatively, the Summary Table (S) contains the start time, end time, number of accepts, and number of rejects for each time a series of master table 312 modifications are provided.
MTCP Generator and Counters
The Generators 308 include two different external counters and two internal counters used in effecting 100% auditability. The two external counters are the Accept Sequence Number Generator and the Reject Sequence Number Generator. The two internal counters are the Total Records Read Counter and the Number of Modifications Counter. All are used only in the Current database 300, as the Archive database 304 is read-only in nature.
Regarding the external counters, the Accept Sequence Number Generator included in the Current database 300 automatically generates sequential numbers for the processing period (daily, weekly, monthly, etc.) starting with the number 1, and increments by 1, so that every transaction processed against the preceding (old) master table 312 will receive a specific transaction number, and accordingly, each transaction processed will be uniquely identifiable based on master table identity, transaction date, transaction number, and authorized user. Note that the transaction date is read off the internal system clock. The Reject Sequence Number Generator counts the number of rejects for the specific processing period. Its function is similar to the Accept Sequence Number Generator. Both the Accept Sequence Number Counter and the Reject Sequence Number Counter are “processing period” specific. That is, both are cleared to zero at, e.g., midnight on the end of the processing period so that each processing period may be separately identified and audited.
Regarding the internal counters, the Total Records Read Counter counts the number of transactions read during a specific processing performance. Since the Total Records Read Counter is “job execution” dependent, this counter is cleared to zero at the outset of every processing program execution. The Number of Modifications Counter counts the number of times a specific record has been changed. As this counter is “record” dependent, this counter is never cleared to zero, This specific counter should identify the number of individual records that may be retrieved, viewed, and
To achieve 100% auditability of a complete system, every master file (or table in relational database management systems has a Master Transaction Cluster. Therefore, a total system containing 15 tables would require 15×8 or 120 tables to achieve full 100% auditability. Since each table will require at least 4 SQL scripts to (1) Create Table, (2) Select data from the table, (3) Delete data from the table, and (4) Drop the Table in the event of redefinition, the number of SQL scripts is 15×8×4, or 960 SQL Scripts. Then, each Master Transaction Cluster will require at least a Processing Program plus a Review, Reset, and Retest, or at least four more programs for each cluster, or 4×15, or 60, more SQL Scripts. All of the SQL scripts would be stored in one SQL Script Library on the computer for future reference and ease of maintenance.
MTCP Multi-processing
The multi-processing of the Master Transaction Cluster occurs in the following manner:
For additions (or Insertions in SQL) of data
-
- The Insertions to the Master Table 312 and
- Insertions to the Accept Table 328 may be processed simultaneously.
For changes (or Updates in SQL) of data
-
- The Update of the Master Table 312 and the Insert to the Accept Table 328 may be processed simultaneously after the original record from the Master Table 312 has been copied to the History Table 332.
For deletes (or Deletes in SQL) of data
-
- The Deletion from the Master Table 312 and the Insertion to the Accept Table 328 may be processed simultaneously after the current record in the Master Table 312 has been updated for the transaction identifier and then copied to the History Table 332.
MTCP Creation
- The Deletion from the Master Table 312 and the Insertion to the Accept Table 328 may be processed simultaneously after the current record in the Master Table 312 has been updated for the transaction identifier and then copied to the History Table 332.
Before processing any Master Transaction Cluster, the necessary databases and files (or tables) must be created. For each business enterprise utilizing the present invention, these databases and files are created only once in the following manner:
MTCP Processing
Processing of the “Master Transaction Cluster” then occurs in the following manner.
- Step 1: All required information for processing a transaction is first captured on an Input Form.
- Step 2: Once this information is edited by, e.g., an operator, an Enter Key can be pressed by an operator to write this information to the Input Table 316 for particular master transaction clusters.
- Step 3: For each input table 316, a polling program notes that the Input Table is not empty and has a transaction action to be processed whereupon the action is processed by a process (denoted “process 1” in FIG. M1).
- Step 4: The transaction processing program determines the type of file maintenance to perform; basically,
- (1) add a record (entitled Insert a Row in SQL),
- (2) change a record (entitled Update a Row in SQL), and
- (3) delete a record (entitled Delete a Row in SQL),
which in turn determines the multi-processing potential as described above in the MTCP Multi-processing.
The normal daily processing flow to achieve 100% auditability in either real-time or batch mode is as follows:
- Step 5: At the end of the “proofing period”, such as daily or weekly, when proof tallies are matched to computer tallies, the Accept Table can be deleted as follows:
- (Begin Program)
- Delete All Records from the Accept Table
- (End Program)
- Step 6: Backup all databases and tables before any information is purged as follows:
- (Begin Program)
- Write All Tables in the “Current” database to backup
- Write All Tables in the “Archive” database to backup
- (End of Program)
Step 7: At the end of a user-defined period, an archive and purge process occurs that
- Step 8: In the event that current records are wrongfully moved to the History Archive,
- they may be retrieved by
- (Begin Program)
- Move Specific Records from the Master Archive to the Master Table
- Move Specific Records from the History Archive to the History Table
- (End Program)
This program should be executed only after Records have been moved from the Current database 300 to the Archive database 304. It should never be run after new transactions have been processed to the Current database 300.
MTCP Backup/Recovery
- they may be retrieved by
If necessary, a recovery program can be utilized at any time in the event of hardware failure. Upon complete recovery, Step 7 and Step 8 will have to be re-executed to insure the correct status before the next day's processing is begun. The Accept Table can then be used to as a substitute Input Table to return the system to its previous processing point. Once this table is exhausted, data from the Input Table would supply the remaining data for the processing job.
MTCP Management
Once test data are defined and processed, a business enterprise may
-
- (a) Review lists of the contents of all Master Tables 312 for determining correctness.
- (b) Reset the contents of all Master Tables for performing the next test.
- (c) Retest.
MTCP Auditability
Once auditabilty is achieved, the business enterprise may query:
-
- (a) When a Master Table Cluster was created.
- (b) When each record was added (or inserted) to the Master Table 312,
- (c) How many authorized changes (or updates) have been made to a record of the Master Table 312.
- (d) Prove the integrity of the master transaction cluster by producing a sequential list of all record changes, and if the record was deleted, where the record is stored.
Accordingly, 100% auditability of every change, every day, for every application is possible.
Multiprocessing Defined
Unlike serial processing which processes all jobs in sequential fashion, multiprocessing processes some of the same jobs simultaneously, or in parallel. While multiprocessing is not new, major computer manufacturers such as Compaq, Digital, Hewlett-Packard, IBM, NCR, Unisys, etc. have announced offerings of low-cost multiprocessing machines based on 2, 4, 8, and sixteen processors. These machines will rapidly increase the demand for multiprocessing software, which is known as “multithreaded” software. Multithreaded software permits the simultaneous execution of more than one jobs or job sequences.
Multiprocessing takes two forms, Symmetrical Multiprocessing (SMP) and Massively Parallel Processing (MPP), the difference being that symmetrical multiprocessing machines collectively have only one bus between the processors and the peripheral storage. For example, a symmetrical multiprocessing machine may have eight processors, one bus, and sixteen disk drives. In contrast, massive parallel processing machines has one bus for each processor. For example, a massively parallel machine may have eight processor, eight busses, and sixteen disk drives. Therefore, symmetrical multiprocessing machines are best suited for applications with a high processing content and a low input/out content. In contrast, massively parallel processing machines are best suited for applications that can be parallelized and have a high input/output requirement, as is the case with many commercial systems.
In either event, multiprocessing machines are best utilized when carefully tuned to avoid bottlenecks. This is likely to mean that all of the layers constituting a computing environment are multiprocessing-enabled. That is, the hardware, operating system, relational database management system, and the specific application are capable of multiprocessing. Some multiprocessing mainframes have been available for several years as well as some versions of the UNIX operating system. Only a few multiprocessing relational databases exist and even fewer multiprocessing applications. It is believed by some that the success of multiprocessing is solely dependent upon the “knowledge of the application” rather than “knowledge of the underlying tools,” the tools being the hardware, operating system, and relational database system.
Accordingly, it is believed that the limiting factors for the success of multiprocessing for financial systems depends on:
-
- (1) the lack of financial transaction application knowledge,
- (2) a lack of understanding of how multiprocessing can be used to effect 100% auditability, and
The value of MTCP is that it addresses the last form of multiprocessing which is believed to be the most critical to delivering rapid response times for real-time financial transaction processing systems. That is, by dividing a transaction into subtransactions that can be spread across several multiprocessors, processing throughput may be faster. Plus, the large number of small programs make maintenance much easier and less expensive.
A first embodiment of the transaction processing controller 52 is provided in the flowchart of
A first embodiment of the transaction preprocessor and decomposer 54 is provided in the flowcharts of FIGS. 7-A through 7-D and FIGS. 8-A and 8-B. Note that for simplicity, error handling and related validity check steps have been omitted. However, the performance of such steps is within the scope of the present invention, as one skilled in the art will appreciate.
A second pseudo-code embodiment of the transaction preprocessor and decomposer 54 follows.
Pseudo-Code for the Edit Processor for all Incoming Transactions (Transaction Preprocessor and Decomposer 54)
Note: Leave all switches=1 until the last routine is completed. This forces the processing to loop through each succeeding routine until completed. Then turn set all swtiches=0 so that the Scheduler will revert back to the Command Program to read another transaction.
A first embodiment of the processing for the subtransaction processing module 64 is provided in the flowcharts of FIGS. 9-A through 9-B,
A second pseudo-code embodiment of the transaction processing controller 52 follows.
Pseudo-Code for Processing for the Subtransaction Processing Module 64
A first embodiment of the processing for the balance sheet table 130 is provided in the flowchart of Fig. BAL-SHT. Note that for simplicity, error handling and related validity checking steps have been omitted. However, the performance of such steps is within the scope of the present invention, as one skilled in the art will appreciate.
A second pseudo-code embodiment of the processing for the balance sheet table 130 follows.
Balance Sheet Processing (BS)
- Note: do not turn switch OFF or back to 0 as these swithces indicate which processes remain.
Note: Do not turn switch OFF or back to 0 as these switches indicate which processes remain.
Notes:
*denotes Primary Control Tables
**denotes “Driving Tables”
The specific invention is a standardized file naming convention to be used in the automatic generation of program code for multiple large-scale transaction processing applications (such as securities trading, telecommunications billing, and work management) on multi-processing computers (using 4, 8, 16, 32 processors) with 100% auditability of user-defined controls. The standardized file naming convention is totally independent of any specific
a.) application such as accounts receivable, customer billing, etc.,
b.) industry such as financial services, telecommunications, or work management,
c.) hardware manufacturer such as Compaq, Digital, HP, IBM, NCR, Unisys,
d.) operating system such as MS-DOS, UNIX, OpenVMS, MVS, etc.,
e.) relational database management system such as Oracle, Sybase, MS-SQL Server,
f.) computer language such as SQL, COBOL, Fortran, PL/1, etc.
The standard naming convention contains the fewest number of characters in any naming conventions; namely, eleven characters used by MS-DOS. The naming convention of MS-DOS uses eight characters as a file name and three characters as a file extension wherein the user may define a file name using the alphabet and selected other characters. While this flexibility is suitable for home use are a small number of files and users, it is not acceptable for large-scale enterprise-wide applications with large number of files and large number of supporting technicians. Hence, the need for enterprise-wide standards.
The standard file naming convent n contains six elements that permit the technician to readily identify the functionality of the specific script (or program) without looking at its contents. Using ANSI Standard structured Query Language as an example language, the six elements are:
a.) a 2-character mnemonic for the SQL commands such as:
b.) a 2-character mnemonic for the application name such as
c.) a 2-character mnemonic for the table (or file name) such as
d.) a 1-character mnemonic for the table cluster role such as
e.) a 1-character mnemonic for the table cluster type such as
- f.) a 3-character extension is then added to the file name depending upon
the type of operating system being used such as MS-DOS, UNIX, OpenVMS, etc. and
whether or not the file is a source file for programmer use or a compiled file (or stored procedure) for machine use.
Hence, script name examples are:
CTXBMDMM.SQL—Create Table for the External Billing System, Master Definition Table Cluster, Master Table, and Master Role for SQL use.
DTXBDCOJ.SQL—Drop Table for the External Billing System, Detail Call Cluster, Output Table, and Journal Role for SQL use.
Circumstances Leading to the Invention
The circumstances leading to the invention of a standard SQL script naming convention are:
a.) one programmer will rarely adhere to the same naming conventions over time and unless an acceptable standard is defined each succeeding programmer added to the job will only complicate the issue by bringing their own standards. Hence, software maintenance becomes a matter of knowing which programmer wrote which program at what time.
b.) without a naming standard any programmer has no idea of what functions the programming is performing without opening the program and examining the program code. This process produces create inefficient maintenance by existing programmers and inefficient training for new programmers.
c.) Competitive pressures are mounting for the efficient of software maintenance.
Advantage of the Invention
Because no duplicate script names are permitted the name of each SQL Script should
a.) convey to the user the precise use of each SQL Script and
b.) permit the storage of all SQL scripts in a one SQL Script Library, or directory.
A standard naming convention also permits the user to determine what scripts may be automatically executed in sequence by use of a SQL command script, which is a single SQL script containing a list of SQL scripts to be executed in sequence. Hence, any single SQL scripts contained in the SQL Library can be reused in many different SQL command scripts.
Although any standard naming convention represents a unique entity separate and apart from the other technologies described immediately above, this particular naming convention is unique in that it embraces all of the logical information necessary to readily identify the role of the script in the total system.
Detailed Description of Invention:
std_name is a standard naming convention that constructs names for programs (or SQL Scripts), system tables, table clusters, and data elements. The seven basic elements are:
std_name defines both “external” names used by the operating system and “internal” names used by the specific program.
where the SQL Script Names are used by the operating systems.
The “internal” resulting names are:
where the Data Element Names are used by the programs (or SQL Scripts).
External Names used by the operating system in identifying programs (or SQL Scripts) are created by employing the following naming components:
Internal Names used by the program (or SQL Script) in processing the data elements are created by employing the following naming components:
Data Tracing
By addressing both the external names for the operating system and the internal names for a specific program, the naming convention is global in nature. In the event that one data element derives its source of input from another table rather than its own specific input screen, then the data name is extended by placing the table identifier of the table supplying the data between the first four and second four characters of the intended data name. Should the data be derived from another table that also derived its data from another table, then eight characters are placed between the first four characters and the last four characters of the intended data name. In the fashion, the data name points backwards through all of the preceding tables to the original source of data and its input form. This process is called “data tracing”, and it provides benefits to programmers in the testing and debugging stages of software development by identifying the original source of data. Thus, “data tracing” provides the programmer with thorough documentation of the data flow throughout an entire system.
Standard naming conventions do not apply to certain language extensions such as the script footings that, for example, specify the size of the table to be created in a “Create Table” script.
The foregoing discussion of the invention has been presented for purposes of illustration and description. Further, comments and description is not intended to limit the invention to the form disclosed herein. Consequently, variation and modification commensurate with the above teachings, and within the skill and knowledge of the relevant art, are within the scope of the present invention. The embodiment described herein above is further intended to explain the best mode presently known of practicing the invention and to enable others skilled in the art to utilize the invention as such, or in other embodiments, and with the various modifications required by their particular application or uses of the invention. It is intended that the appended claims be construed to include alternative embodiments to the extent permitted by the prior art.
Claims
1-26. (canceled)
27. A method for processing a financial transaction on a computing system using a plurality of processors comprising:
- retrieving, in response to a request to perform a financial transaction having a transaction identifier and input values, a unique subset of user-defined algorithms from a set of user-defined algorithms, wherein the unique subset is associated with the transaction identifier via a user-defined transaction processing table, each user-defined algorithm, when executed by a processor on the input values, creates or changes a data record such that the unique subset of user-defined algorithms associated with the transaction identifier, when executed using the input values, creates or changes data records stored in a database as necessary to process the financial transaction;
- executing each of the user-defined algorithms of the unique subset on the input values;
- wherein the unique subset of algorithms are executed concurrently on different processors and the financial transaction is considered processed when all algorithms in the unique subset of user-defined algorithms have been executed on the input values.
28. The method of claim 27, wherein each request further includes a licensee identifier and the set of user-defined algorithms is further associated with the licensee identifier.
29. The method of claim 27, wherein executing each of the user-defined algorithms of the unique subset on the input values comprises:
- executing a user-defined algorithm that creates a data record in a transaction journal file;
- executing a user-defined algorithm that changes a data record in a general ledger master table;
- executing a user-defined algorithm that changes a data record in an account master table; and
- executing a user-defined algorithm that changes a data record in an entity attribute master table.
30. The method of claim 27 further comprising:
- generating a transaction sequence number for the retrieved request to perform a financial transaction.
31. The method of claim 29, wherein executing each of the user-defined algorithms of the unique subset on the input values comprises:
- executing a user-defined algorithm that creates, changes or deletes a data record in a cash flow table;
- executing a user-defined algorithm creates, changes or deletes a data record in an income statement table;
- executing a user-defined algorithm that creates, changes or deletes a data record in a capital gains table;
- executing a user-defined algorithm that creates, changes or deletes a data record in a balance sheet table; and
- executing a user-defined algorithm that creates, changes or deletes a data record in a performance measurement table.
32. The method of claim 27, further comprising:
- if the transaction identify is not associated with a unique subset of user-defined algorithms from the set of user-defined algorithms, writing the request to perform the financial transaction to a system reject table.
33. The method of claim 27 further comprising:
- if one of the algorithms fails to execute, reversing all algorithms of the unique subset that did execute; and writing the request to perform the financial transaction to a system reject table.
34. The method of claim 27 wherein each of the unique subset of algorithms is executed concurrently on a different one of the plurality of processors.
35. The method of claim 27 wherein at least one of the unique subset of algorithms is executed on a first processor concurrently with the execution of at least one algorithm of a different unique subset of algorithms on a different processor.
36. The method of claim 27 wherein at least one of the unique subset of algorithms is executed on a first processor concurrently with the execution of at least two algorithms of different unique subsets of algorithms on different processors.
37. A method for processing financial transactions on a multiprocessing machine having a plurality of processors, each financial transaction having financial transaction data, the method comprising:
- maintaining a set of user-defined algorithms, each user-defined algorithm being independently and simultaneously processable by any one of the plurality of processors in the multiprocessing machine and each user-defined algorithm when processed on financial data adds, changes or deletes only one financial data record based on the financial data;
- receiving a first financial transaction having a first user-defined financial transaction type and first financial data;
- identifying, for the first financial transaction, a first unique subset of user-defined algorithms from a set of user-defined algorithms based on the first user-defined financial transaction type;
- processing each user-defined algorithm in the first unique subset of user-defined algorithms on the first financial data;
- receiving a second financial transaction having a second user-defined financial transaction type and second financial data;
- identifying, for the second financial transaction, a second unique subset of user-defined algorithms from the set of user-defined algorithms based on the second user-defined financial transaction type;
- processing each user-defined algorithm in the second unique subset of user-defined algorithms on the second financial data; and
- wherein all the user-defined algorithms in the first unique subset of user-defined algorithms on the first financial data and at least one of the second unique subset of user-defined algorithms are concurrently processed.
38. The method of claim 37, wherein each user-defined algorithm changes only one financial data record in only one user-defined financial report and includes an operator, a first operand that identifies input data from the financial transaction data, and a second operand that identifies the only one data record.
39. The method of claim 37 further comprising:
- maintaining a user-defined transaction processing table that associates each user-defined transaction type with a unique subset of the set of user-defined algorithms including associating the first unique subset of user-defined algorithms with the first user-defined financial transaction type and the second unique subset of user-defined algorithms with the second user-defined financial transaction type.
40. The method of claim 37, wherein processing each unique subset of the set of user-defined algorithms associated with a user-defined transaction type comprises:
- executing a user-defined algorithm on a first processor that creates a data record in a transaction journal file;
- concurrently executing a user-defined algorithm on a second processor that changes a data record in a general ledger master table;
- concurrently executing a user-defined algorithm on a third processor that changes a data record in an account master table; and
- concurrently executing a user-defined algorithm on a fourth processor that changes a data record in an entity attribute master table.
41. The method of claim 37 further comprising:
- wherein the at least one of the second unique subset of user-defined algorithms is completed prior to completion of processing all the user-defined algorithms in the first unique subset of user-defined algorithms.
42. The method of claim 37 further comprising:
- receiving the second financial transaction after receiving the first financial transaction; and
- completing processing of all of the user-defined algorithms in the second unique subset of user-defined algorithms prior to completing the processing of all of the user-defined algorithms in the first unique subset of user-defined algorithms.
43. A financial transaction multiprocessing system for an enterprise, each financial transaction having a financial transaction type and financial transaction data, comprising:
- a multiprocessing computer having a plurality of processors;
- a financial record database storing financial data records;
- a database manager for creating, changing and deleting financial data records;
- a queue for receiving financial transactions for processing, wherein each financial transaction when processed results in adding, changing or deleting to at least one financial data record and wherein each financial transaction is associated with one of a plurality of user-defined financial transaction types and includes financial transaction data;
- an algorithm database storing a set of user-defined algorithms, each user-defined algorithm capable of being processed concurrently and independently with any other user-defined algorithm and each user-defined algorithm when processed by any one of the plurality of processors causing the database manager to add, change or delete a financial data record; and
- a transaction processing table that associates each user-defined transaction type with a unique subset of the set of user-defined algorithms such that processing a first financial transaction having a first financial transaction type and first financial transaction data is achieved by processing a first unique subset of user-defined algorithms associated with the first financial transaction type on the first financial transaction data.
44. The financial transaction multiprocessing system of claim 43 further comprising:
- a set of control tables;
- a set of user-defined control algorithms, each user-defined control algorithm when processed causing the database manager to add, change or delete one or more data records in one or more control tables; and
- wherein the transaction processing table further associates one or more user-defined control algorithms with each user-defined financial transaction type.
45. The system of claim 43 further comprising:
- a processor queue for user-defined algorithms, the processor queue distributing queued user-defined algorithms to processors as the processors become available.
46. The system of claim 43, wherein each user-defined algorithm includes only one operator and only a first operand and a second operand, the first operand identifying input data from the financial transaction data of the financial transaction to be processed, and the second operand identifying the only one data record.
Type: Application
Filed: Aug 26, 2004
Publication Date: Jul 28, 2005
Inventor: William Hinkle (Denver, CO)
Application Number: 10/928,463
