Consistent Hashing Ring illustration

Consistent Hashing Ring

Draws the consistent hashing ring used by memcached clients, Cassandra and DynamoDB: servers are hashed onto a circle of 2³² positions, and every key belongs to the first server clockwise from it. Add and remove servers and the tool highlights exactly which keys moved — roughly 1/N of them — while a side-by-side "hash mod N" comparison shows the naive approach remapping almost everything on the same change. A virtual-nodes slider demonstrates the second half of the trick: one point per server makes a lumpy, unfair ring, but 20–50 virtual nodes per server smooth the load out, which you can verify in the live load histogram and the stats table (keys per node, % of keys, % of ring). Type any key and watch the lookup walk clockwise to its owner; switch to rendezvous (HRW) or Maglev hashing to compare the tradeoffs; and copy a scenario link that restores your exact nodes, vnode count, key count and algorithm.

Runs 100% in your browser — nothing you paste leaves your device.

Notes

  • With plain "hash mod N", going from 4 to 5 servers remaps about 80% of all keys — every cache behind it goes cold at once.
  • Consistent hashing moves only the keys that fall between the new node and its predecessor: about K/N keys, no matter how many servers exist.
  • Virtual nodes fix load imbalance: a single point per server can leave one server owning half the ring; dozens of points per server even it out statistically.
  • Amazon's Dynamo paper popularized this exact design; the ring metaphor survives in Cassandra's "token ring" terminology.
  • Runs 100% in your browser — nothing you paste leaves your device.