Hacker News Reader

Comment by saryant | Hacker News Reader

Excellent post. We made the mistake of trying to use Chef to deal with seed node discovery in our Akka cluster and that was a terrible idea. Really wish we'd gone with the AWS SDK instead.

View on HN

Thanks! The SDK has some "quirks" (and isn't especially Scala-friendly) but I reckon it's ultimately pretty powerful. My main gripe is the lack of real-world examples in the docs - hence my desire to write this up.

One question: how do you deal with crashed JVMs? You can't rejoin a new JVM if it has the same address and port but I can't think of a good way to communicate the new port using the AWS SDK.

The way we have it set up, if the JVM crashes then it will mean that the health check url used by the load balancer will become unresponsive. This would lead to the auto scaling group downing the instance and replacing it with a new one. So, the same node wouldn't ever need to rejoin.

Hi, I didn't see an obvious way to reach you on the blog bit just FYI, it looks weird on iPhone Safari

Edit http://imgur.com/oX0CzuZ http://imgur.com/WsqEVxA

Hi, thanks for the feedback! Yeah I'm aware the look on Chrome on my Android is pretty terrible, which is on my list to get sorted. Looks like it's even worse on iOS! At the moment its basically using the default site template from Jekyll which is apparently very much not mobile first!

Also re: contact, there's a link to my Twitter account but not much more. Cheers for letting me know anyway!

Edit: Grokked media queries; hopefully a little nicer on mobile now.

Were you using Chef inside of OpsWorks? I'm curious about what problems you're having.

I'd think that hooking the configure lifecycle event on your cluster layer in OpsWorks would work just fine.

No, just Chef Solo. We only had four nodes at the time.

Reply on news.ycombinator.com