Prometheus 2 is out

https://prometheus.io/blog/2017/11/08/announcing-prometheus-...

For google cloud, stackdriver is usually sufficient for most queries

https://cloud.google.com/monitoring/

We use a combination of splunk, solarwinds' server & application monitoring and new relic. We're about to start using prometheus for monitoring some apps that we've moved into containers. We also have several thousand monitors/alerts that are setup in ipmonitor that we're trying to get moved over to sam.

For just logs, LogDNA is fast and cheap.

For everything else, Datadog has the best setup although their agent is old and messy, but they have logging coming soon and some basic APM integration.

If kubernetes, try Sysdig or Netsil.

Datadog's agent v6 was rewritten in Go recently:

https://www.youtube.com/watch?v=egSvw7xYw9s

The Datadog agent is well built for K8s, amongst many other things.

If you think the Datadog agent is old I have bad news: Sysdig and Netsil rely on v5.x of it.

Has to be PRISM, but I'm not sure if NSA is sharing access. /s

I am not sure about the best, but we use Zabbix at our company.

I highly recommend checking out Blue Medora products (full disclosure: I work there). We offer world class monitoring products across a lot products and platforms.