,
4 tweets,
1 min read
Read on Twitter
YES THIS
Data at rest is dead data. Immutable data isn't really data at all, or not the data that matters; it's just scaffolding and structure to lend predictability to your real data, which is always in flux and transition.
Data at rest is dead data. Immutable data isn't really data at all, or not the data that matters; it's just scaffolding and structure to lend predictability to your real data, which is always in flux and transition.
And yet we treat it in exactly the reverse: we optimize for stale/dead data, and treat changes like the anomaly.
Your business is an event built on top of events. Shift your frame and so much shit falls neatly into place.
Your business is an event built on top of events. Shift your frame and so much shit falls neatly into place.
Shit like... why are traditional statsd-style metrics *actively undermining* your efforts to understand your systems?
(Because resting state is dead state, all you actually care about about are events and their context. What is happening? What's changing?)
(Because resting state is dead state, all you actually care about about are events and their context. What is happening? What's changing?)
Life gets easier when you look at data at rest as a partially expired event state.
