Zero Downtime Archiving Model for financial applications
Mücahit Gündebahar, Merve Can Kuş Khalilov · 2013
In financial systems, current data reading ratio is greater than the historical data. This study describes an approach to continuous archiving on a defined time interval for active and inactive data within a database. In financial systems, downtime is never desirable in operations such as database maintenance. Database-level archiving operation can be in many ways. However, there is often need for short-term downtime for this operation. In this study, a structure will be proposed which has a downtime occurrence within milliseconds. Archiving can be done to another database on the same server or can be done to another database on a different server. The first reason for archiving is acceleration to data access. The second target is reducing cost for processor, memory, disk and network resources. With archiving, beside data access, maintenance and administration facilitation can be enhanced. At the same time, it has also a simple methodology to access and query the inactive data. The proposed method requires partitioning with certain intervals, daily, monthly or weekly, to be archived. Seamless data transfer can be possible with partitioning for a table on a particular schema and function. And finally, the active and inactive data separated under a partitioned view. In this way, the database management system can explore and know the data location and access during the query processing.