Startup Mistakes: Choice of Datastore
A great advantage of having a large network of technical friends is that they ask you for advice on things, which I love giving. One great disadvantage of people is that they rarely take my advice without justification, even though I think everybody should know better by now. A discussion I frequently have with friends (and which they don’t just blindly take my advice on), is their choice of datastore, which invariably goes something like this:
- Trust me, don’t use MongoDB.
- Why, what’s wrong with it?
- Look, how many times have I given you some advice, you didn’t listen, and later on it turned out I was right?
- Ah, so you’re saying I should use Cassandra.
So, since I keep having to justify my opinion (can you believe that? Just ridiculous.), I figured I’d do it once, in this post, and then I can just point people here when they’re about to do something dumb. If I linked you to this article, this means you.
EDIT: Apparently the self-deprecating sarcasm above wasn’t really very obvious, and it comes off as arrogant, but my intention was for it to be satire (cleary opinions should be justified, even mine). Also, the Cassandra joke was a reference to this lady. Like an ancient Greek proverb says, “the best joke is one you have to explain on your blog”.
Datastores are important
The datastore is often the most important part of