Compute Optimisation - methods of reducing cloud costs
I will be writing a series of brief insights, in which I will discuss in detail, the methods of consumption cost reduction. Best suited to cloud consumption ( AWS or Azure or Google or other ), to achieve lower cloud opex spend, but also equivalently applicable to datacentre infrastructure use, where successful compute consolidation and efficiencies would reduce capex spend - where yearly infrastructure increases to match business growth could be reduced or stemmed entirely.
The focus areas to be addressed will fall into the following approaches, which if executed successfully, will generate large savings and substantial reduction in spend associated with compute, database and storage :-
Scheduling - run it only when you need it
Demand scaling - run more when transaction demand is high and run less other times
Compute reservations - forecast consumption demand at discount
Database reservations - lock in long term database computation to gain discounts
Right Sizing - assessing current load against the actual runtime execution performance
Enterprise Discount Plans - multiyear commitment to lock in a savings percentage
Latest versions - leveraging Mores Law of ever improving technologies over time
Cloud Native - server less
Stay tuned as I publish in the coming days each of the approaches.
