System Design and Scalability Interview Questions | Qdrant | NextRound

1What is the relationship between shard count and query fan-out cost, and why doesn't 'more shards' always mean 'faster'?Advanced

2Design a semantic search system that must support 500 million documents with sub-100ms p99 latency. What are the key architectural decisions?Advanced

3How would you plan capacity (RAM, disk, CPU, node count) for a collection of a given size, vector dimensionality, and expected QPS?Advanced

4What architectural changes would you make to support near-real-time search over data that changes thousands of times per second (e.g., a live feed)?Advanced

5How would you design a system that needs to support both 'search the last 24 hours' and 'search all history' with very different latency expectations?Advanced

6What role does caching play in a Qdrant-backed search system, and at what layers would you introduce it?Advanced

7How would you decide the initial number of shards for a new collection when the eventual data size is uncertain?Advanced

8How many replicas would you configure for a shard serving a mission-critical, read-heavy workload, and what does each additional replica cost you?Advanced

9What operational steps are involved in adding a new node to an existing Qdrant cluster and rebalancing shards onto it?Advanced

10How does Qdrant's architecture and target use case differ from Pinecone's as a fully managed, closed-source vector database?Intermediate

11When would you choose pgvector inside an existing Postgres database over a dedicated vector database like Qdrant?Advanced

12What distinguishes Qdrant from Weaviate and Milvus at a conceptual level, and what would make you choose one over the others for a given project?Advanced

13Under what circumstances would a team be justified in NOT using a vector database at all, and instead using brute-force search or a traditional search engine?Advanced

14How would you validate that a newly restored cluster from snapshots is actually healthy and serving correct results before routing production traffic to it?Advanced

15What is the operational difference between a rolling upgrade of a replicated cluster and an in-place upgrade of a single-node deployment?Advanced

16How would you design a disaster-recovery strategy that survives the loss of an entire cloud region?Advanced

17What is your target Recovery Point Objective (RPO) and Recovery Time Objective (RTO) for a Qdrant deployment, and how do snapshot frequency and replication factor influence each?Advanced