Detect app/request bottlenecks and trace them

[abitrolly]

#119 is about detecting OS load/failures and tracing their source independent of application that caused it. The issue here approaches the problem of degraded performance/failures from application side, by using Application Performance Management techniques, such as tracing, logging, alerting by metrics.

Tools to consider:

• https://zipkin.io/
• https://github.com/etsy/statsd

[hleb-albau]

https://prometheus.io/ + https://grafana.com/
But it is not story for cybernode, its application story(ies).

[abitrolly]

@hleb-albau how do you see tracing our API search timeout or slow response time with Prometheus + Grafana?