Day 54/50 Rendezvous Hashing va Consistent Hashing in Load balancing algo

Hinzugefügt: 2026-05-25

[00:00:00]One concept of hashing which 90% of developers would fail to answer in a tech interview. Here we go with the problem. Design a distributed cache that minimizes remapping when the server goes down. And most of you will be saying that consistent hashing is the right approach for it. But do you know there's an alternative solution named rendezvous mapping which is equally efficient when it comes to load balancing. Hello everyone, welcome to day 54 of system design interview preparation series and today we're going to talk about rendezvous hashing. How does it operate?

[00:00:24]Why is it so efficient? And what are the pros and cons of it over consistent hashing? So here we go. Consider you have a client library which they're implemented in the load balancer. It operates through hash function and these are the given set of nodes node A, B, C, D. What it will do, it will compute weight for the incoming key for each of the node and the highest weighted value is a clear winner. Here node B is a clear winner for the key K and key K will reside in node B. Now you must be wondering what happens when a node gets deleted from the entire system by the load balancer. All the keys that resides within node A will be recomputed and distributed among the remaining nodes.

[00:00:55]The keys that exist between B and C still are the clear winners by virtue of the highest value being assigned in them. When a node is added to the entire block, for example A, B, C exist and D is getting added, all the keys that exist in A and B and C will be recomputed and compared with the value D. If D is a clear winner, then those will be shifted, otherwise they remain intact. For rest of the details, the coding algorithm, what are the pros and cons over consistent hashing, everything is mentioned in the blog. Do check the description.