Tyler Treat

Yep, that's pretty much exactly what I was thinking. Also allowing daemons to register themselves so the client doesn't necessarily have to specify peers. For example, maybe I start a...

Are you seeing the coordinator as something that would go in place of etcd? It'd be easy enough to roll something simple since it basically just needs to listen for...

Yeah, that sounds good to me.

Google loves to break their oauth2 lib... On Fri, Jan 16, 2015, 1:20 AM savorywatt [email protected] wrote: > Initial pass here > https://github.com/savorywatt/Flotilla/tree/etcd > > Ran into some flotilla-server compilation...

Yeah, many thanks to @codahale for pinging me re: tinystat and usl because they look really interesting/applicable. Right now, the daemons only send back to the client a few data...

Isn't the [`Cleaner`](https://github.com/travisjeffery/jocko/blob/e2a8d10c648b5a558711dcfa20e166c4ad2d7c73/commitlog/delete_cleaner.go#L3) interface responsible for compacting/truncating the log for purposes of preventing unbounded disk usage, i.e. `DeleteCleaner`? It appears this is what uses `MaxLogBytes` to control the log size:...

So it appears one issue is this does not actually seem to do what it says: https://github.com/travisjeffery/jocko/blob/1c69c5c274caf8d6b3a7953c67e5de51c55378be/commitlog/index.go#L200-L208 The issue is that the memory-mapped index file is 10MB, so reading at...

I have it working with a binary search. Will submit a PR shortly.

@travisjeffery One snag I've run into: I'm using a fork of the commitlog package which I've modified index entries to also contain a `Size` field which is the size of...

Would it be feasible to automate it?