Community Award Winners at CouchConf San Francisco
CouchConf San Francisco was a total blast and we at Couchbase want to shout out a huge thanks to those of you who joined in. I had a great time, and one of the highlights of my day was getting to announce the Couchbase Developer Award winners. And the winners are (envelope please):
Best in Show — Group Complete by Radical Dynamics (submitted by Matt Adams)
Group Complete is a collaborative and real-time mobile data collection solution. Using Group Complete’s Inform app for Android, mobile workers can build electronic “smart forms,” distribute them to their team members and share completed forms with the entire group. Group Complete takes advantage of CouchDB’s ability to transparently replicate data and operate while offline allowing team members to take their data wherever they need it and ensure that changes are seamlessly shared with their mobile co-workers as soon as a network becomes available.
Also big thanks to the Group Complete team on their community contributions to the Android efforts.
Couchbase Mobile — CouchPad (submitted by Marty Schoch)
Couchpad is a general-purpose CouchDB utility designed for the iPad. Couchpad gives the user a touch interface to all of the most common CouchDB operations. Users can browse and edit their documents, explore views and set up replications. Couchpad even has the ability to detect Couchapps deployed in the database and enables users to take them full-screen in an embedded browser. Further, since Couchpad is built on the Mobile Couchbase iOS platform, it includes an embedded CouchDB instance. This means a user can browse their CouchDB’s in the cloud, then use replication to pull the data down to their iPad and work with it locally. Couchpad also has the ability to advertise itself via Bonjour and find other nearby instances. This makes Couchpad ideal for the next replication party!
Couchbase/CouchDB — Blitz.io by MuDynamics (submitted by Kowsik Guruswamy)
Blitz.io’s architecture is described in detail here. Blitz.io uses two regional CouchDB clusters with multi-master replication. The scale engines that generate the load tests are distributed around the world and listen in on the _changes feed to a CouchDB that’s geographically closest. When a user runs a load test, they use filtered changes to “wake up” the appropriate regional scale engines and then use the conflict mechanism to acquire the job. The scale engines have zero local storage and completely rely on the update_seq in the CouchDB. This allows them to spin up and auto scale these engines simply by booting up the AMI and the whole thing looks and feels like an IRC channel.
Kowsik is also an awesome member of the CouchDB community, thanks for everything that you contribute!
Membase Server — PlayTime Platform by TubeMogul (submitted by Nicolas Brousse)
Membase Server is used in TubeMogul’s Real-Time Bidding infrastructure. They handle over 30,000 qps which is multi-billions ads requests per day on which we are bidding. Each bid request must be responded in less than 120ms including network round trip. With such constraints, every part of their solution needs to be scalable, robust and fast. Their 95th percentile Membase Server response time is under 2ms on EC2. They also have over 15 Membase servers spread around the world.
Our next category was for the Pillow of the Community award, and that went to Max Ogden, a Code for America Fellow, who is spreading the word about Couchbase all over, with projects like Recline and PDX API. Thanks Max for all that you do!
For those of you who missed the event, we have good news! We announced that we will host CouchConf Berlin on November 7 and will be announcing a few more locations in the near future. Stay tuned!
Congratulations once again to our award winners!