[mikeryan]

This is weird to me.

"We started thinking seriously about deploying Cassandra in production around three weeks ago. After looking at the site for something that would be a good fit, we settled on green badges."

It seems completely baffling to me that someone would go out and compare different db solutions, pick one and THEN try to find a way to fit it on their site architecture.

[jbellis]

I read that as, "which feature is easiest to port from the existing mysql solution?" Which is a reasonable question to ask, given how different the two databases are.

[mikeryan]

My process would be this.

1. Decide that mysql (or other RDB) is problematic.

2. Decide which features cause the most pain (List them with the most painful on top)

3. Discover which alternative db solutions would do the best in alleviating the top issues.

[jbellis]

4. Start implementing, starting with the most-improvement-for-least-effort.

I'm saying the two lists are not mutually exclusive. :)