This post is a framework to contain my experience in attempting to leverage higher performance in BAU tasks under a rapidly increasing data load.
In my environment, we are entering a phase of increased data granularity. Leveraging the ability to receive & process interval data from meter service providers.
Legacy business processes capable of supporting the low impact requirements for Monthly invoices from register readings have grown to the point were we are on the cusp of invoicing customers on Half Hour data.
During this step change in data granularity we are noticing what worked before is insufficient going forward. I intend to document my findings from a functional perspective with a minor focus on the technical aspects underpinning reading a quarter of a million records a month, to tens of millions of records per day.
IS-U-DM-MR
Meter Reading Upload
Production runtime in seconds
EXADATA runtime in seconds
HANA runtime in seconds
IS-U-BI
Billing & Invoicing
Production Runtime in seconds
- Billing 2300
- Invoicing 650
- Print 5600
EXADATA runtime in seconds
- Billing 800
- Invoicing 380
- Print 1700
HANA runtime in seconds
