Lecture
Modern databases are among the most widespread specific objects in the field of informatization, for which in a number of domains
especially high quality and its qualified systematic design are required. A database can be viewed as two components:
the software of the database management system (DBMS), independent of the field of application, structure and semantic content of the data being accumulated and processed;
the database information (DBI), available for accumulation, ordering, processing and use in a specific problem-oriented field of application.
At the same time, one and the same database management system (DBMS) can process data that differ in structure, composition and content, and the same data can be managed by the software of different DBMSs. Although these components interact closely in the implementation of a specific applied database, initially, at the design stage, they are created or selected practically independently and can be regarded, over their life cycle, as two objects, which differ in:
the range and content of the quality indicators that determine their purpose, functions and consumer properties;
the technology and automation tools for developing and supporting the whole life cycle of each object;
the categories of specialists responsible for: creating, operating or applying the database components;
the sets of operational and technological documentation supporting the objects' life cycle.
The first component for systematic analysis and quality requirements is the DBMS software complex. Practically the entire set of software quality characteristics and attributes set out in the ISO 9126 standard can, to one degree or another, be used when forming quality requirements for a DBMS. In all cases, the most important quality characteristics of a DBMS are the requirements for functional suitability for the processes of forming and modifying the database's information content by administrators, as well as for data access and the presentation of results to database users. Below, as a basis, the range and content of the standardized characteristics of complex software systems are adopted, adapted to the concepts and features of database components. Depending on the specific problem-oriented field of DBMS application, priority in the systematic analysis of quality requirements may be given to various constructive characteristics: either reliability and security of use (the financial sphere), or ease of use for low-skilled users (the social sphere), or efficient use of resources (the logistics and supply sphere). However, in practically all cases some role is retained for a number of other constructive quality indicators.
The second component of a database is the information itself that is actually accumulated and processed. In database systems, the data themselves, their storage and processing, acquire dominant importance. Below, emphasis is placed on the systematic analysis of the requirements and constituent quality characteristics of this object — on the database information (DBI) with the assumption that the DBMS tools are able to provide them. For evaluating the quality of database information, the same general methodological approach can be retained for identifying the appropriate range of characteristics standardized in ISO 9126 basic characteristics and sub-characteristics of software quality. The quality indicators identified should be of practical interest to database users and should be ordered in accordance with the priorities of practical application. In addition, each quality indicator identified for the DBI must be suitable for sufficiently reliable evaluation or measurement, as well as for comparison with the required value during customer testing.
When designing each database, the contract, technical specification and specification must select and formalize a representative set of functional requirements for DBI quality, adequate to its purpose and field of application, as well as to the requirements of the customer and potential users. Just as for software, the quality characteristics of the DBI can be divided into functional and constructive. Their range, content and sub-characteristics below are based on the descriptions recommended by the ISO 9126 standard. They appear to be sufficiently universal and applicable for systematizing the quality characteristics of database information. However, the range of quality indicators cannot always be limited only to the characteristics
of the information in the database, and must include a number of refinements reflecting the overall effectiveness and functional suitability of the joint use of the DBMS and the DBI by users under real conditions.
Functional suitability of the DBI can be a complex problem when determining how well requirements correspond to the real values of the necessary quality attributes, especially for large distributed databases when using diverse and complex information about the objects being analyzed. A measure of the quality of functional suitability can be the degree of coverage of the database's goals, purpose and functions by the information available to users. Just as for software, for databases within functional suitability it is advisable to use a group of sub-characteristics defining the functional and structural requirements for databases. In addition, the functional suitability of many DBIs can be reflected by:
the completeness of the accumulated descriptions of objects — the relative number of objects or documents present in the database, to the total number of objects on the given subject, or in relation to the number of objects in similar databases of the same purpose;
the identity of the data — the relative number of object descriptions free of defects and errors, to the total number of documents about objects in the DBI;
the timeliness of the data — the relative number of outdated data about objects in the DBI, to the total amount of accumulated and processed data.
The constructive characteristics of database information quality can, on the whole, be considered to include, with some refinement of the concepts, sub-characteristics and attributes, practically all the standardized software quality indicators presented in ISO 9126. Requirements for database information must also cover ensuring its reliability, efficient use of computer resources, usability — applicability, maintainability™ and portability. The content and attributes of these constructive characteristics in this case differ somewhat from those applied to programs, but their essence, as basic concepts and quality characteristics of objects, is advisable to use during design for the systematization and regulated formation of requirements for these
system components. The measures and scales for evaluating constructive characteristics can, to a large extent, be the same as those used when analyzing the quality of software.
Correctness or reliability of data — is the degree to which information about objects in the database corresponds to the real objects outside the computer at a given moment in time, determined by changes in the objects themselves, inaccuracies in the records of their state, or inaccuracies in the calculation of their characteristics. During systematic design, the selection and establishment of requirements for data correctness in the database can be evaluated by the degree of coverage by accumulated, up-to-date and reliable data of the state and changes of the external objects that they reflect (see Table 11.1). In addition, correctness of the database can also include some volume-time characteristics of the stored and processed data:
the volume of the database — the relative number of records of object or document descriptions in the database, available for storage and processing, compared with the total number of real objects in the external environment;
responsiveness — the degree to which the dynamics of change of data descriptions during collection and processing correspond to the states of real objects, or the amount of permissible delay between the appearance or change of a real object's characteristics and its reflection in the database;
depth of retrospective coverage — the maximum time interval from the release date and/or the recording in the database of the earliest document up to the present time;
dynamism — the relative number of modified object descriptions to the total number of records in the database over a certain time interval, determined by the periodicity of the release of database versions.
Security of database information is mainly implemented by the DBMS software, but in combination with the supporting data organization and protection tools. The goals, purpose and functions of protection are closely related to the functional-suitability features of each DBI. When designing, the property of protecting database information from negative impacts is usually described by the composition and range of methods and tools used for protection against external and internal threats.
Reliability of database information can be based on applying the concepts and methods of reliability theory, which makes it possible to obtain a number of clear, measurable integral indicators of their quality. A reliable DBI must, first of all, ensure a sufficiently low probability of loss of operability — a failure, during its real-time operation. A rapid response to the loss or distortion of data, and the restoration of its reliability and operability within a time shorter than the threshold between a fault and a failure, ensures a high reliability of the database. If, in these situations, sufficiently rapid recovery occurs such that no failure is recorded, then such events do not affect the main reliability indicators — the mean time between failures and the availability factor of the DBI. The unpredictability of the type, location and time of occurrence of DBI defects during operation makes it necessary to create special, additional systems for the operational protection against unintentional, random data distortions. Reliability must be increased through means of ensuring noise immunity, operational monitoring and recovery of the DBI.
The ISO 9126 standard recommends analyzing and taking into account the reliability of software complexes through four sub-characteristics, which can also be applied for forming requirements for the quality characteristics of database information. Completeness — a property of the DBI consisting in its ability to avoid falling into failure states as a result of losses, distortions, errors and defects in the data. Stability to defects and errors — is a property of the DBI to automatically maintain a specified level of data quality in the event that defects and errors occur, or the established interface for data exchange with the external environment is violated. For this purpose, it is recommended to introduce into the DBI the operational detection of information defects and errors, their identification, and the automatic restoration (restart) of normal DBI functioning.
Recoverability — is a property of the DBI, in the event of a failure, to restore the required level of information quality, and also to correct damaged data. This requires computational resources and time to detect the inoperable state, diagnose the causes of the failure, and implement the recovery processes. Availability or readiness — is a property of the DBI to be fully able to per-
form the required function at a given moment in time under specified conditions for the use of the database information. The generalization of failure and recovery characteristics is carried out in the criterion of the DBI availability factor. This indicator reflects the probability of having recoverable data in an operable state at an arbitrary moment in time.
Efficiency of resource utilization of the computer, when analyzing the real functioning of the DB, is reflected by the time characteristics of interaction between end users and DBI administrators in the process of operating the database for its direct purpose. Temporal efficiency of the DB is determined by the duration of performing specified functions and waiting for results from the DBI in average and/or worst cases, taking into account task priorities. It depends on the volume, structure and speed of data processing, which directly affect the time interval for completing a specific computational process, and on throughput — performance, i.e., on the number of jobs that can be performed on the given computer within a specified time interval (see Table 11.2).
Resource utilization or resource economy in the standards is reflected by the occupancy of central processor resources, main, external and virtual memory, input-output channels, terminals and communication network channels. This value is determined by the structure, functions and volume of the DBI, as well as by the architectural features and available resources of the computer. Depending on the specific tasks and features of the DBI and the computer, when designing and selecting DBI quality attributes, either the absolute value of occupancy of resources of various types, or the relative value of resource utilization of each type during normal DBI functioning, may dominate.
Practicality — usability — often significantly determines the functional suitability and usefulness of applying the DBI for qualified users. The users may include administrators, end users and indirect users, who are influenced by or depend on the quality of the database information. This group of quality indicators includes subcharacteristics and attributes that, from various angles, reflect the functional clarity, ease of learning, system efficiency and simplicity of data use. Some subcharacteristics can be evaluated using economic indicators — the labor and time costs of specialists for implementing certain data-interaction functions (see Table 11.3).
Clarity depends on the quality of the documentation and the subjective impressions of potential users regarding the functions and characteristics of the DBI. In the project, it can be qualitatively represented by the clarity of the functional concept, the breadth of demonstration capabilities, the completeness, comprehensiveness and clarity of presentation in the operational documentation of the possible functions and features of data implementation in the DB. It must be ensured by the correctness and completeness of the description of the source and resulting information, as well as of all the details of using the DBI for users.
Ease of use of the DBI— the ability to operate it and manage data conveniently and comfortably. This requires ensuring: a sufficient set of control parameters implemented by default, informativeness of messages to users, clarity and unification of screen control, as well as the availability of changing DBI functions in accordance with user qualifications, and a minimum of operations required to carry out a given task and analyze the results. Some attributes of this subcharacteristic are available when setting quantitative requirements by specifying the labor input and duration of the corresponding processes of training and instructing qualified users for the effective operation of the DBI.
Learnability can be defined by requirements for the labor input and duration of preparing a user for the full-fledged operation of the database information. The learnability of the DBI depends on the internal properties and complexity of the structure of the DB information, as well as on the subjective qualification characteristics of specific users. It can also be characterized by the volume of operational documentation and/or the volume and quality of electronic textbooks.
Maintainability of the DB information in the project can be reflected by the convenience and effectiveness of correcting, improving or adapting the structure and content of data descriptions depending on changes in the external application environment, as well as in the requirements and functional specifications of the customer. In general, the quality of DBI maintainability can be represented by the need for labor and time resources
to ensure and implement it. The possible expenditure of economic, labor and time resources for the development and improvement of DBI quality depends not only on the internal properties of the data, but also on the requests and needs of users for new functions, and on the readiness of the customer and developer to satisfy these needs. Based on the volume of the anticipated changes, as well as the data newly introduced in the next version, taking into account the complexity and novelty of their development, requirements for their implementation can be formulated.
The set of subcharacteristics of software maintainability, as presented in the standard ISO 9126, is quite applicable for describing the requirements for this quality indicator of database information, largely through the same organizational-technological subcharacteristics. Analyzability of the DBI depends on the coherence of the architecture, the unification of interfaces, and the completeness and correctness of the technological and operational documentation for the DB. Changeability consists in the adaptability of the structure and content of the data for implementing specified changes and extensions and for data configuration management. Changeability depends not only on the internal properties of the DBI, but also on the organization and tool support of the maintenance and configuration management processes, to which the architecture and the external and internal data interfaces are oriented in the project.
Testability depends on the magnitude of the scope of the changes that must be tested when structural and content modifications occur in the DBI data, and on the complexity of the tests needed to verify their characteristics. Its attributes depend on the clarity of formalization, in the systems project, of the rules for the structural construction of the components and of the whole DBI complex, on the unification of inter-module and external interfaces, and on the completeness and correctness of the technological documentation. The subcharacteristics of changeability and testability of data are amenable to quantitative determination by the magnitude of the labor input and duration of implementing these functions for typical data operations when applying various automation methods and tools.
Mobility of the DB data, just as for programs, can mainly be characterized by the duration and labor input of their installation, adaptation and replaceability when porting the DBI to other hardware and operating platforms. Information about the processes occurring in the external environment may have a large volume and labor input for initial accumulation and updating, which determines the need for its careful storage and regulated modification. Since porting a DB is often driven by the need to increase the computer resources available for solving new, promising tasks, their project becomes a natural extension of the DBI's functions relative to the original version of the project. To assess quality and determine requirements for DBI mobility, just as for software, it is necessary to solve the problem of comparing the achieved effect and costs for methods of porting or redeveloping the components and populating the database in specific conditions, taking into account all the listed factors and costs.
Comments