Reading peak traffic without drowning in charts
How to separate meaningful workload shifts from noisy spikes when your application faces a known busy period.
Field notes
Guides from the work
Articles rooted in application analytics—how we read workloads, judge scaling posture, and learn after capacity incidents.
What a scaling policy cannot see
Autoscalers react to the metrics you give them. Here is how to spot the blind spots that leave capacity behind the curve.
Choosing an observation window for a workload study
Too short and you catch noise. Too long and you blur the patterns that matter. A practical way to set the window.
After a capacity incident: a calmer debrief agenda
A structured debrief keeps blame out of the room and puts workload facts back in the center.