What is a Hash Function? What makes a good hash function?

0-2 years experience

1. 1Suppose you need to store usernames in a small in‑memory hash table. How would you pick a hash function, and what would happen if many usernames hash to the same bucket?
2. 2If you use Java's String.hashCode() for keys in a HashMap and notice a sudden slowdown after adding a few entries, what could be causing it and how would you investigate?
3. 3When implementing a simple hash table from scratch, how do you handle collisions, and why does the choice of hash function matter for that?

2-5 years experience

1. 1We shard user data across 64 partitions using a hash of the user ID, but some partitions are hot while others are idle. Walk me through how you would diagnose if the hash function is the problem and what alternatives you might consider.
2. 2Our caching layer uses a custom hash function for keys, and after a recent change the cache hit rate dropped dramatically. How would you debug this and decide whether to replace the hash function?
3. 3When designing a Bloom filter for duplicate detection, how does the quality of the hash functions affect the false‑positive rate, and what criteria would you use to pick them?

5-8 years experience

1. 1Design a distributed hash table for a key‑value store handling billions of keys with low latency. What properties must the hash function have, and how would you mitigate uneven key distribution or hot spots?
2. 2Our microservice uses consistent hashing to route requests to backend nodes. After a node failure we see a latency spike. Explain how the hash function choice impacts load redistribution and what you could change to reduce the impact.
3. 3If you need to migrate from a legacy hash function to a new one without downtime, what strategy would you use to keep data consistent and minimize disruption?

8+ years experience

1. 1Our company is consolidating several services that each use different hash functions for sharding data across a shared storage cluster. How would you standardize the hashing strategy across teams while handling backward compatibility, operational risk, and future scalability?
2. 2We are moving from a monolithic hash‑based partitioning scheme to a multi‑tenant, geo‑distributed architecture. What high‑level design considerations around hash function selection, determinism, and cross‑region consistency would you raise?
3. 3When building a platform that exposes a pluggable hashing API for third‑party developers, what safeguards and design patterns would you enforce to prevent poor hash functions from degrading overall system performance?