NETOBSERV-122: Health stats reporter

[mariomac]

Status report based on Eclipse Microprofile 2.1.

Implements /health/ready and /health/live endpoints, returning something like:

{
    "checks": [
        {
            "data": {
                "host": "goflow-kube-54485b9d94-s695x"
            },
            "name": "flows",
            "status": "UP"
        }
    ],
    "status": "UP"
}

It also implements a /metrics endpoint to know the exact information about flows processed, templates received, etc...:

# HELP flow_process_nf_count NetFlows processed.
# TYPE flow_process_nf_count counter
flow_process_nf_count{router="100.64.0.2",version="10"} 229
flow_process_nf_count{router="100.64.0.3",version="10"} 22
flow_process_nf_count{router="100.64.0.4",version="10"} 39
flow_process_nf_count{router="100.64.0.5",version="10"} 103
# HELP flow_process_nf_delay_summary_seconds NetFlows time difference between time of flow and processing.
# TYPE flow_process_nf_delay_summary_seconds summary
flow_process_nf_delay_summary_seconds{router="100.64.0.2",version="10",quantile="0.5"} 59
flow_process_nf_delay_summary_seconds{router="100.64.0.2",version="10",quantile="0.9"} 60
flow_process_nf_delay_summary_seconds{router="100.64.0.2",version="10",quantile="0.99"} 1.8446744073709552e+19
flow_process_nf_delay_summary_seconds_sum{router="100.64.0.2",version="10"} 2.9514790517935283e+20
(....) etc

[jpinsonneau]

That's smart !

But does it mean if there is no flows coming for a time (typically hanging at r.format.Next()) you will not push boxes ?

[mariomac]

You are right, but it should not matter since in the stats aggregation we just count the boxes whose start time is less than 5 minutes.

[jotak]

we said dromedary :)

[mariomac]

ouch! Actually this struct is not converted to JSON anymore. Removing the json tags

[jotak]

I haven't checked in deep, but does it capture when flows are discarded due to not having the definitions? Maybe it would requires some change directly in goflow, as I'm not sure if we are notified about that at our level.

[mariomac]

@jotak mmm... good point. I was assuming that the goflow library was returning error on that situation. I'll double-check

[jotak]

@mariomac we also need a PR in the operator, to have goflow-kube service expose these endpoint, right?
never mind, i've just seen it :)

[jotak]

FYI I created https://issues.redhat.com/browse/NETOBSERV-168 as a follow-up, to provide the full status as first intended

[jotak]

/lgtm

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

Approval requirements bypassed by manually added approval.

This pull-request has been approved by:

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment