Performance Always Matters
What do I mean by that? Well, specifically, what I mean is that the performance of the application code that you write, dear reader, always matters.
There is no point in the powers that be funding a go-faster analytic platform if we can’t be trusted to feed it performant code.
I’ve seen far too many instances of multi-million pound systems brought to their knees by poor SQL code.
As well as delivering the required business logic, it’s also important to carry out performance testing prior to implementation.
That code might work fine on the dev dataset you’ve used, but how will it perform in production at 10x or 100x the scale used in dev?
In addition to not wanting to cripple the expensive analytic platform, it is also important to implement performant code to ensure we don’t run out of processing slots in the schedule.
Rest assured, as data volumes, data sources and demand for data products increase, the day will remain firmly stuck at 24 hours in length.
Performant code ensures we can do as much as possible with the compute resources at our disposal, rather than having to buy our way of trouble with a costly and avoidable platform upgrade.
Write neat code, write correct code and write performant code. That’s the law!
#data #sql #performance #tuning #analytics #performanceengineering #performancetuning #speedking