Interview with Apache CouchDB Project Leaders on Features, Use Cases, and Future Directions

Welcome to the "Deep Dive into Databases" series, featuring an interview with Apache CouchDB Project Management Committee members Adam Kocoloski and Jan Lehnardt, where they share insights on CouchDB's advantages, disadvantages, roadmap, and expert advice for running CouchDB on Kubernetes.

Background: Adam works on IBM Cloud's database and data services strategy, while Jan has been involved in open‑source CouchDB for over a decade, contributing to CouchDB 3.0/4.0 and the Opservator monitoring tool.

How they joined CouchDB: Both were attracted by the database's focus on data distribution, security, and durability, seeing an opportunity to apply their physics‑based data handling experience to a non‑traditional database.

Strengths: CouchDB’s unique replication (similar to Git) supports point‑to‑point, multi‑region, and cluster‑wide synchronization, offering strong data durability and ease of use; its REST API is praised for simplicity and long‑term usefulness.

Weaknesses: Achieving low‑latency, out‑of‑the‑box performance can be challenging, especially due to TLS handshake overhead and limited client‑library support, which can steepen the learning curve.

Cool use cases: CouchDB powers IBM Cloud services, mobile games, digital wallets, global logistics, in‑flight entertainment systems, Ebola outbreak response tools, and vaccine trial software, demonstrating its suitability for distributed, offline‑first applications.

Future expectations (CouchDB 3.0 & 4.0): 3.0 will address common user complaints, while 4.0 will introduce a FoundationDB‑based distributed transactional key‑value store, improving consistency and query performance by running view indexes on FoundationDB.

Kubernetes deployment advice: Deploying CouchDB in containers and Kubernetes offers better automation and scalability, but teams should have dedicated resources for cluster management; the goal is an out‑of‑the‑box experience that works across cloud providers.

Advice for newcomers to Erlang: Learning Erlang is not required; contributors can start with JavaScript, Elixir, or the HTTP JSON API, and the community encourages diverse contributions and provides support for getting productive quickly.