Node color = its stored value Yellow flash = write reconciled Live link Severed link
⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.

CAP Theorem Cluster Simulator

Distributed databases replicate data across many nodes so that no single machine failure loses information, but a network partition forces a hard choice: stay consistent by refusing writes the rest of the cluster cannot see, or stay available by accepting them everywhere and letting the copies drift apart. This simulator makes that trade-off visible. A ring of replica nodes shares one value; triggering a partition splits it into two isolated groups with their cross-links severed, and every simulated client write then either gets rejected (CP mode, enforcing consistency) or accepted locally on both sides (AP mode, enforcing availability) — watch the diverged-node counter climb in AP mode, or the rejected-write counter climb in CP mode, and see how healing the partition reconciles the cluster back to one shared value.