Results after couple of good days of coding over Easter: next @bitfinex core platform upgrade will bring an important reduction of latency in the internal queues and core processes, resulting also in around 3x/5x increased capacity in events/sec.
-
3 days to code = 3x capacity. Is this scalable?
It took a while to collect enough prod data and formalize it in possible changes. The changes were actually quite easy to implement due to our modular structure.
