Host metrics - mac

[daniel-flex]

Need host metrics for mac - currently unavailable.

https://lighthouse-book.sigmaprime.io/api-vc-endpoints.html#get-lighthousehealth

[paulhauner]

On it!

[paulhauner]

Making good progress here: sigp/lighthouse#1848

I'll come back to it tomorrow morning to finish it off.

[paulhauner]

I've added this feature here: sigp/lighthouse#1848. @daniel-flex I have sent you a binary privately.

The output looks like this:

    "cpu_status": {
      "status": "ok",
      "message": "CPU below 85%",
      "gauge_pct": 0.12
    },
    "memory_status": {
      "status": "ok",
      "message": "1 GB available memory",
      "gauge_pct": 40.7
    },
    "eth1_status": {
      "status": "error",
      "message": "Eth1 sync is disabled, use the --eth1 CLI flag to enable. Eth1 is only required for validators.",
      "gauge_pct": 0
    },
    "p2p_status": {
      "status": "warn",
      "message": "Low peer count (3).",
      "gauge_pct": 5.45
    },
    "database_status": {
      "status": "ok",
      "message": "/home has sufficient capacity.",
      "gauge_pct": 64.53
    }

[daniel-flex]

@paulhauner have connected this up the the frontend in commit #32ab08d. It's currently coming from the BN api: /lighthouse/health (I was previously getting it from the VC api).

Most of the data is available, but I think this endpoint could be updated slightly, like adding the beacon node health/status info, plus potentially merging/combining some of the fields (as I'm currently doing it on the frontend right now). Can discuss further if you want - I have some ideas about it.

Also, polling timeout can be controlled via an env variable to give the user more control - if the lighthouse client is local then polling could be faster, and if remote then slower, for example. Defaults to 5000ms.