Home Insights Banking Group achieves major Summit database savings with PuriPhi

Banking Group achieves major Summit database savings with PuriPhi

News

PuriPhi, our industry leading online data archiving solution for Summit, is now live for the international investment banking arm of one of the top banking groups in France.

In the words of the Summit system owner:

“PuriPhi is on the way to take away one third of our current data and reunite ten years of archived database snapshots. It allows us to rethink our view of historical data and it will greatly improve our daily testing environment and regular audit requests”

Context: Our client has a 1.6Tb Summit database, replicated across 30 environments for production, test and development. The existing infrastructure was unable to support the future growth of the database. Previous purging of the production database had resulted in multiple historical snapshot copies being maintained, with an inability to easily combine the data for reporting.

Goal: To reduce the live database size and present all data (live, newly archived AND previously deleted data) in a single Summit environment, compatible with data model updates. Maintain the live database at a reduced size according to specific archiving requirements.

Solution: PuriPhi was installed and configured to archive matured and cancelled trades and historical market data more than 415 days old. Even without considering any of the other data types that PuriPhi supports, 500Gb of data was removed from the production database. The PuriPhi solution presented the archived and production data, seamlessly combined with no loss of functionality through the usual and familiar Summit GUI.

Benefits:

– Live database is reduced by 500Gb
– Space savings are multiplied over 30 environments
– Performance of the live system is improved
– Infrastructure supports future growth of database
– Increased number of dev/test environments supported
– Improved dump/load times
– Aggregated view of ALL data for the first time
– Compatible with future data model changes