After two years in production, the enterprise IT monitoring system we wrote for German insurer PBeaKK crossed a major milestone with over 1 billion database records.
Dubbed "IT Cube", this real-time monitoring application runs in a moderately equipped Virtual Machine with a Windows OS, 4 CPU cores and 36 GB RAM, connecting to a vanilla Microsoft SQL Server. Over 3,000 IT entities report state information every 5 minutes.
Codemart's project lead Călin commented:
"I'm shocked, SHOCKED! I never thought our software would have to deal with over a billion records. I was already impressed when it handled a few million fast and well."
Says Dr. Schöpf from PBeaKK:
"Das ist der Hammer!"
Even at such a giant and still growing size, IT Cube remains perfectly responsive with its snazzy user interface, unimpressed by queries reaching deep down into its silo of events. Achieving such a feat without high-end hardware resources is made possible by a clever data model: basically, IT Cube is an OLTP server which behaves like an OLAP system, based on a Star Schema.