Siraaj Khandkar

Consider accumulating StatsD messages in the process mailbox, rather than concatenating lists

1

With per-process metrics added, all these concatenations may get expensive. Will see how it looks in the next few days...

The output can be a stream of edges, like: ``` host,time,parent,child ``` which can then be grouped by time to create graph snapshots at given times, which then can even...

Blows up Graphite disk usage when table names are dynamically generated. Bad practice, but encountered in the wild, so have to deal with it...

Just like in the StatsD one

Ostensibly: ``` erlang -spec produce() -> A. -spec package(A) -> B. -spec transport(B) -> {}. ```