Built in Kibana alerts for APM Server

Currently, **Stack Monitoring** alerts on Kibana give us inbuilt rules for monitoring Elasticsearch and node health. These alerts enable proactive monitoring for Elasticsearch by having rules around shard size, thread pool, etc. We should have similar inbuilt alerting rules to monitor APM Server health and status. For example:

1. High memory usage by APM-Server
2. High CPU usage by APM-Server
3. CPU throttling of APM-Server
4. Elasticsearch output efficiency of APM-Server (% of events sent to ES over total events processed by APM-Server)