Provide benchmark reports

[qinxiongzhou]

This project is awesome! Can you provide benchmark report?

[xzchaoo]

Hi @qinxiongzhou . Sorry for the incomplete documentation.
We'll add the benchmark reports as soon as possible.

Here is an internal real deployment scene. You can refer to this example.

1. The cluster has 40 nodes, about 800 pods, about 800 log-monitoring configs.
2. For each node, we deploy a holoinsight agent and cadvisor daemonset.
   • cadvisor (limit 0.25C 256Mi ) daily CPU 50%
   • holoinsight-agent (limit 0.25C 256Mi ) daily CPU 5%
3. The server side has 2 instances of holoinsight-server, whose specification is 4C8G, with daily cpu_util < 10%.
4. With an internal metrics storage system which consumes 3 x 4C16G.
5. The cluster generates 500,000,000 points daily