Data are inevitable part of human life. In the past, only small data portions were processed, however, nowadays, data significance and relevance is high, it is necessary to manage huge data amount, data requirements are still rising. Adding temporal data definition extends actual conventional paradigm of processing and brings real problems based on performance degradation. In this paper, we propose column level temporal architecture, which is useful for data with various frequency and time of the attribute changes. It requires sophisticated access methods for data retrieval, therefore index structures are proposed. Adding unique and reverse flag, performance results manifested by processing time can be compared. In the experiment section, impact of temporal indexing for data manipulation statement type is described, compared and evaluated.
展开▼