ddev-learn
Sign inSign up

Practice judgment, not trivia.

Every day, work through a realistic scenario from your world. Pick how you'd handle it, see how every option actually plays out, then talk it through with an AI discussant before you save what you learned. Not a course. Not a quiz.

Sign upAlready have an account?
networkingobservabilitysenior

Intermittent 502s only from one mobile carrier

You're on call. Pages start firing at 7:42 AM. The alert: HTTP 502 rate is 12% on /api/v1/auth/refresh, but only for one carrier in one region. Your PM Slack-DMs you: “is this our problem or theirs?”

AReply to PM: "Investigating, will update in 15." Then dig into logs filtered by carrier ASN.
Positive
Sets expectations, buys time, and signals you're not flying blind.
Negative
15 minutes can vanish; if you miss the deadline you erode trust faster than not committing.
DReply to PM: "It's the carrier — their problem," based on the regional pattern. Monitor.Common pitfall
Negative
Diagnosing before any data is the fastest way to be confidently wrong in the postmortem.
Common pitfall
A green carrier status page is not evidence. 502 means your edge returned it — the request reached you.

Every option gets revealed — including the two you didn't pick.

Built for how you actually make decisions

Scenarios and the AI discussion match your role — the tradeoffs a product manager weighs aren't the ones an incident commander weighs.

Software engineer

Incident response, architecture, and technical trade-offs

Product manager

Scope, timeline, and stakeholder pressure calls

Incident commander

Coordination and comms while an incident is live

Engineering manager

People, process, and team judgment calls

Ready to sharpen your judgment?

Sign up