Twitter tells all

Here's a disarmingly honest, yet embarassing, post about Twitter's recent brokenness from their official blog.

Choice quote:

Q: Is it true that you only have a single master MySQL server running replication to two slaves, and the architecture doesn’t auto-switch to a hot backup when the master goes down?
A: We currently use one database for writes with multiple slaves for read queries. As many know, replication of MySQL is no easy task, so we've brought in MySQL experts to help us with that immediately. We've also ordered new machines and failover infrastructure to handle emergencies.

Right, okay, it's not super-easy. I mean, your average person who writes crap websites by copying and pasting bits of PHP probably couldn't do it. But, erm, if you're running a quite popular service with stupid amounts of funding, it's probably something you should be able to manage.

And that's without even mentioning partitioning. I really don't understand what their issue is.

The post in question doesn't exactly give me much hope for Twitter's future, quite frankly, especially now that people like Google and Facebook, who both maintain vastly larger IM systems with less trouble, are muscling in on their territory. I bet the VCs who funded the thing are cringing right now...