> 2. These distributed databases like Cassandra, MongoDB, and CouchDB[3] aren't actually very scalable or stable. Twitter apparently has been trying to move from MySQL to Cassandra for over a year. When someone reports using one of these systems as their primary data store for over 1000 machines for over a year, I'll reconsider my opinion on this.
Are there facts to back up what he says about MongoDB for example or is this just FUD.
His assertion is actually backed up by MongoDB docs: "Sharding will be production-ready in MongoDB v1.6, estimated to be released in July 2010." Their current sharding alpha is only tested/supported for small clusters (up to 20 nodes). In addition, looking at their sharding mode of operations, only a very limited set of actions are properly sharded; the rest require touching all the shards.
Disclaimer: this was the first look I've had at MongoDB. However, based on this alone, it would be prudent to wait several years before considering using this technology in serious production. It is somewhat reckless to consider a DBMS feature that wasn't locked for awhile "production-ready".
Are there facts to back up what he says about MongoDB for example or is this just FUD.