Comment by Nina Okafor on "Which vector DB actually scales without becoming a maintenance nightmare?"

CommentWhich vector DB actually scales without becoming a maintenance nightmare?

Nina Okafor

ML engineer working on LLMs and RAG pipelines

Postgres with pgvector honestly works better than people expect if you're not doing millions of QPS. We run 40M vectors in prod and reindexing happens in the background without downtime using IVFFlat + HNSW hybrid approach. Pinecone pricing is brutal once you start scaling.

Milvus consistency is solid if you actually read the docs on consistency levels, but K8s overhead isn't worth it unless you're already running that stack. For 50M vectors at typical RAG scale, Postgres + pgvector saves you from operational complexity that usually isn't necessary.

Alex Petrov

Systems programmer. Rust evangelist.

2d ago

Postgres+pgvector is solid for moderate scale, but HNSW reindexing without downtime is a lie if you're using standard Postgres transactions. You need careful planning around index replacement. That said, self-hosted beats Pinecone's SaaS tax for sure once you own ops overhead.

pgvector is solid for most use cases. the hybrid approach makes sense if you're optimizing for both recall and latency. though once you hit scale constraints, the operational overhead of managing postgres itself can bite you harder than a dedicated vector db. depends if that's worth the cost savings.

Search Hashnode