Pass the CKA by drilling decisions, not topics

Start here

1. Read the curriculum — the exact domains and weights, transcribed from the CNCF PDF, with the date they were fetched.
2. Read the retrospectives — eight recurring themes from public pass/fail write-ups. Almost every reported failure is a process failure.
3. Read the strategy page, then never practise untimed again.
4. Set a target date and let the scheduler decide your daily workload.

The domains

How the study tools work

• Review session — flashcards and cloze drills through an SM-2 scheduler. Ease factor, interval and repetition count per item; overdue first, then new. Fully keyboard-driven.
• Practice tasks — scenario-based and timed, each with a model solution and a verification command. The timer turns red when you pass the target, because that is the moment to flag and move on.
• Mnemonics — for what must be recalled cold: the kubeadm upgrade order, etcd’s mandatory TLS flags, taint effects, the three conditions that make a PVC bind.
• Dashboard — per-domain mastery and time spent, sorted by exam risk (curriculum weight × how far you still are from mastery), so the weakest high-weight domain is always first.
• Study plan — target date in, daily workload out, with the last three days reserved for review only.

Everything is stored in localStorage. No backend, no account, nothing leaves your browser — so export your progress if you would mind losing it.

Content policy: no exam dumps

The CKA is under NDA. Nothing here scrapes, reproduces, or approximates real exam questions, and no such material was consulted while building it.

What the practice tasks are: original scenarios inferred from two public sources — the published curriculum weights and competency list, and community retrospectives about preparation and process. If a task pattern looks high-yield here, it is because the curriculum weights it heavily and people repeatedly report underestimating it, not because anyone reported seeing it.

Community write-ups are cited by URL in the retrospectives, along with an honest note about which pages could be fetched directly and which were read through search summaries.

Offline

The one-book reference PDF is generated from the same markdown and JSON as this site — cheat sheet, every decision tree, every mnemonic, and the flags worth memorising. It regenerates on every build, so it cannot drift from the site.