Atef R.

Assistance (scoping and implementation) for database migration projects in private cloud

• Assistance (scoping and implementation) for database obsolescence management projects

• Definition of architecture and provision of databases for new applications

• Handling of project support requests for databases

• Optimization requests

• Database Design (advice on modeling, index positioning, ...)

• Audit, upon request

Expertise:

- Study DBA roles

- PostgreSQL

English-speaking environment

The mission In this context, it is transforming and working on the adoption of new private/public Cloud solutions for its database infrastructure. The service will take place within the Study DBA entity, which is responsible for project support for PostgreSQL, Oracle, and MongoDB technologies on Linux and AIX.

I was involved in two rather complicated projects, as DBAs had been working on optimization for 2 years without satisfactory results:

Project for DSN Purge, according to the General Data Protection Regulation (GDPR), personal data cannot be kept indefinitely. Monthly Nominative Social Declarations (DSN) from companies containing personal data will be kept for a maximum of 5 years.

The objective is to increase the deletion speed from 173,575 lines/min to 372,222 lines/min. By applying adequate tuning techniques (advantage of partitioning), I managed to reach a deletion speed of 987,179 lines/min, which means a performance gain of X3 compared to the objective.

Project 2: Optimization of the DS4 Overhaul Program. The objective is to optimize the program for checking the conformity of DSN data received against the contract parameterization data.

There, I managed to reduce processing time from 23 hours to 30 minutes by adding missing indexes for the slowest and most time-consuming queries, and by updating the database statistics or metadata.

In March 2022, I joined FIS as a Senior Analyst DBA, joining the UBIX development team. My main role consisted of analyzing and resolving anomalies in UBIX, a global back-office platform for derivative product management. I also played an important role by offering advice and technical support to UBIX clients, whether by sending best practices for configuration and developing recommendations for optimal use of the UBIX platform, or by preparing patches for clients and writing resolution procedures to ensure client satisfaction.

For example, I detected issues in the AWR report in the "sga target statistics" section, and I often recommended allocating a minimum of 20% of the total Oracle server memory to ensure adequate PGA optimization.

For query tuning, I use these steps: identify slow queries, then collect necessary information to analyze the causes of slowness, and then propose adequate solutions for each situation (rewriting, indexing, bind variables, plan fixation, purging tables with unnecessary historical data...).

One of my significant achievements was conducting in-depth studies to improve UBIX on Oracle, particularly on the integration of Oracle Advanced Queuing (AQ), thus making UBIX compatible with the RAC architecture. This initiative led to a major evolution, offering a robust and scalable architecture for UBIX to meet the needs of compatibility with the RAC architecture, thereby strengthening its reliability and performance.

Furthermore, I would like to highlight that CASEIS was one of our valued clients during my experience at FIS, where I provided technical support to resolve UBIX performance issues.