I'm wondering if anybody will easily notice me if MongoDB or CouchDB are prepared for any production atmosphere.

I am now searching at these storage solutions (I am favouring MongoDB right now), however, these projects are very youthful therefore i anticipate that I am going to need to work quite tough to convince my manager that people should adopt this new technology.

What Let me know is:

  1. Who's using MongoDB or CouchDB today inside a production atmosphere?

  2. How's it going using MongoDB/CouchDB?

  3. What problems (if any) did you discover whenever you adopted this new storage mechanism (and just how have you overcome them)?

  4. How have you cope with any migration problems that you needed to cope with?

  5. Have you got worthwhileOrpoor encounters with either of those solutions that you might want to talk about?

I am the CTO of 10gen (designers of MongoDB) so I am a bit biased, however i also manage couple of sites which are using MongoDB in production.

businessinsider continues to be using mongo in production for more than a year now. They're utilizing it for from customers and blogs, to each image on the website.

shopwiki is applying it for any couple of things including real-time statistics along with a caching layer. They're doing over 1000 creates per second to a pretty big database.

If put forth the mongodb Production Deployments page you'll many people who're using mongo in production.

For those who have any queries concerning the scale or scope of production deployments, publish on our user list and we'll be more than pleased to assist.

The BBC and meebo.com use CouchDB in production and thus do you of my clients. Here's a listing of others using Couch: CouchDB within the wild

The main challenge would be to understand how to organize your documents and prevent thinking when it comes to relational data.

SourceForge uses MongoDB.

I'm using CouchDB in production. Presently it stores all individuals 'optional' fields that were not within the original DB schema. And right now i'm considering moving all data to CouchDB.

The correct answer is a dangerous step, The truth is that. First of all, since it is not v1. yet. And next, since it is drivespace-hungry. By my information, CouchDB file (with indexes) is ~30 occasions bigger than MySQL database with similar rows. However I think it'll exercise all right.

CouchDB .11 (launched in the finish of March) is really a feature-freeze release for 1.. What this means is we'll be maintaining compatibility using the current API for 1., now is a great time for you to visit again CouchDB should you haven't shortly.

The CouchDB .11 source code release can be obtained here. You will find binary contractors along with other treats linked here.

We use couchdb in production and also have since right before the project went underneath the Apache umbrella.

It's used by us to keep exactly what we may otherwise make use of a dbms, plus a variety of unstructured data. Personally, I like the best way to just throw a variety of data in it and employ the sights to cull what you do not need with respect to the situation.

The toughest part was leaving the dbms mindset. We authored our very own migration utils once the storage format transformed just safe, to ensure that wasn't a real problem.

We've not had any negative encounters yet, however again we've not had the setup under any type of huge load. I think things works pretty much since we now have two slave type servers that replicate from one master server that will get all the creates. I am confident that people do not have to get it done this way for replication to operate properly, but it is the way we arrange it at first also it stuck.