For the programmers: I highly recommend
this recording from a webinar the other week. It is the Deputy CTO of Typesafe explaining -- well, a whole lot of stuff. Rather than being about nuts and bolts, he spends about an hour covering the *rationale* behind the Typesafe stack.
It's really good stuff, especially if you aren't already bought into all of this. He lays out, step by step, how to break away from all the wishful thinking that makes so many systems unreliable, and how the Akka project has gradually evolved a set of tools built around the reality that systems *are* unreliable, how to build a decently reliable cluster anyway, how to deal with scalability, the tradeoffs between latency and throughput, and tons more. It's probably the most concise overview of all of this stuff I've ever seen, and regardless of whether you are going to buy into it, it's worth taking an hour to at least understand the logic.
Besides, how can you resist a talk with a section titled "Resilience and Supervision in the Zombie Apocalypse"?
Check it out, and I strongly encourage asking me questions about it if you have any. This is the bedrock that Querki is built on top of...