Interview: Meet CouchBase
CouchOne and Membase recently announced they will merge to form the new NoSQL company, 'Couchbase' and launch a new product family, which will be based on the Apache CouchDB document database technology. In this interview, JAXenter speaks to co-author of O'Reilly's 'CouchDB: The Definitive Guide,' about the move......
JAXenter: Membase and CouchOne have announced a merger. In the announcement, you stress how well these two companies fit together. What makes Membase and CouchOne such a good fit?
Jan Lehnardt: At a technology level, Membase and CouchOne were starting at different points, but heading in the same direction on our roadmap. Just as an example: Membase was looking to develop indexing and querying, which were strong points of CouchDB. CouchOne heard from users the need for clustering and elasticity, strengths of Membase. And both companies had roadmap plans and engineering activity to precisely fill the holes. Merging fills both our gaps 1-2 years faster than we would have been able to individually, and assembles a much larger team to innovate in non-duplicating directions.
At a company level, there were also great synergies. For example, Membase was looking for a CTO, CouchOne was looking for a CEO as we went through a round of funding. Again, we filled each other's gaps in this area as Damien is the new CTO of the merged company, and Bob Wiederhold, a very experienced CEO, is providing business leadership.
JAXenter: You state your intention to join both products together. What will the resulting product look like?
Jan Lehnardt: We have a high level plan, but the details are being worked out as we speak. A few characteristics are:
- Full API compatibility with Apache CouchDB and Memcached. This allows us to offer a drop-in replacement for users that already use either of the projects.
- The CouchDB persistence, indexing and replication layer adds these features to Membase. CouchDB users will benefit from higher performance and strictly low latency guarantees of Membase.
- Our mantra is "Simple, Fast, Elastic" and it is not just some marketing bullshit. There's barely an engineering discussion where this isn't brought up. We believe in software that is easy to get started with, is easy to use, does not compromise on making the most of current and future hardware and allows users to scale out to massive cluster and millions of mobile devices for data sync. If we find that we are discussing a feature or implementation that would break the above promise, we won't do it.
JAXenter: What will pre-existing customers from each company, gain from the merger?
Jan Lehnardt: First, we'd like to emphasize that for users and customers whose needs are currently met with the status quo, nothing will change except the url of the company website. We are committed to make Couchbase a drop-in replacement for both Apache CouchDB and Membase / Memcached. At the same time, we merged precisely because many of customers were asking us for new features. Membase users keep asking for more advanced querying options, CouchDB has that. CouchDB users ask for higher performance and latency guarantees, Membase has that. These are only two items on a very long list that we gathered from our respective users.