Short, honest write-ups of the ideas that come up in interviews and on the job. Each one teaches the concept properly, cites the Microsoft documentation, and points back to the class that drills it until it sticks.
Premium SSD bills by size tier, not usage — most fleets pay P30 prices for P10 workloads. Measure real IOPS at P95, map to the tier ladder, and know the resize asymmetry before you buy. When Premium SSD v2 ends the problem.
The VNet is free; everything attached to it bills. The billing map for peering, NAT Gateway, private endpoints, gateways and firewalls, how to trace each meter in Cost Analysis, and the three quiet monthly burners.
Compare a complete workload, its free-plan limits, and the cost after the trial. The provider with the largest credit does not automatically have the lowest ongoing bill.
Check your subscription, read every meter, and remove finished resources. A budget sends a warning; it does not switch the lab off.
Reuse your Windows Server and SQL licenses to drop the software fee — the Software Assurance catch, the vCore math, and how to toggle it.
GPU compute is the whole bill. Right-size, spot, reserve, scale to zero — and know when serverless model APIs beat renting GPUs at all.
Both cut the same bill. A reservation says "I will run this exact VM size in this exact region for a year"; a savings plan says "I will spend five dollars an hour on compute" — and the price of that easier promise is a shallower discount.
A budget is the thirty-second insurance policy against a runaway cloud bill — it emails you as actual or forecasted spend climbs toward your threshold. The catch everyone misses: it warns, it does not stop the spending.
Most overspend is not clever waste — it is idle VMs, over-sized everything, and pay-as-you-go prices on workloads that run all year. The levers that move the bill, in order of effort to payoff.
One short, honest Azure note at a time — plus the occasional hiring signal. No spam, no card, unsubscribe in one click.