Oracle In-Memory dramatically improved data processing speeds of Korea SSIS.
By using Oracle's In-Memory features, it was possible to analyze and batch process large amounts of data more than 10 times faster than before applying them.
“Thanks to the Oracle In-Memory feature, we resolved concerns about DB performance deterioration and achieved information system operation efficiency and optimization of some business processes.”
Business Challenges
Korea SSIS performed data analysis and batch processing using stored procedures in the Oracle 11gR2 Database, which took several hours or more. Consequently, processing the entire SSIS welfare services took too much time, making it very difficult to respond effectively when problems arose.
Why Customers SSIS Chose Oracle
Korea SSIS found that the performance degradation of the existing DB system was due to disk I/O limitations when processing large amounts of data, and they wanted to convert their stored procedure batch programs into Java programs for various reasons. However, they were concerned that this change might further decrease performance. To address this issue, they were looking for ways to process data at the DB memory level. In this context, the Oracle In-Memory feature, which allows for direct processing of large amounts of data in DB memory, was the desired DB feature for SSIS. This is why Korea SSIS upgraded their Oracle DB to 19c and utilized the Oracle In-Memory feature.
Results
The problem with the DB system used previously was that as data increased, I/O problems occurred and the data processing speed quickly slowed down. Additionally, as expected, when the stored procedure batch programs were converted into Java programs, a level of performance degradation occurred that rendered them unusable.
To effectively solve this problem, Korea SSIS upgraded Oracle 11g DB to 19c and took advantage of Oracle In-Memory feature.
After loading about 11 important business tables into Oracle memory using the Oracle In-Memory feature, we succeeded in quickly analyzing and batch processing them with the parallel processing.
Additionally, about 11 large tables were effectively loaded into Oracle memory by utilizing the compression function provided by Oracle In-Memory.
Thanks to these Oracle 19c In-Memory features, we were able to convert all stored procedure batch programs into Java programs, and confirm data analysis and batch processing performance that was about 10 times faster than before applying In-Memory.
In conclusion, as DB data processing speed has improved, the overall SSIS welfare work processing time has become much shorter than before.It is possible to respond quickly in the event of a system failure, enabling more stable welfare services to the people.
About the customer
The Social Security Information Service(SSIS) in KOREA is a leading institution of information platforms that helps Koreans conveniently receive welfare benefits. Also, it aims to support the enforcement of KOREA government's welfare policies.