#metrics #statsd #environment #seconds #daemon #numbers #date

app countdown-metrics-rs

A daemon that publishes the number of seconds until a given date to a statsd sink

2 stable releases

1.2.0 Apr 20, 2023
1.1.1 Apr 20, 2023

#12 in #seconds

MIT license

17KB
212 lines

Countdown Metrics

A daemon that publishes the number of seconds until a given date to a statsd sink.

Configuration

The daemon is configured via environment variables:

  • ENV_PREFIX: The prefix for environment variables. Defaults to COUNTDOWNS_.
  • INTERVAL: The number of seconds between publishing metrics. Defaults to 10.
  • STATSD_HOST: The host to publish metrics to. Defaults to 127.0.0.1:8125.
  • METRIC_PREFIX: The prefix to use for metrics. Blank by default.
  • METRIC_TAGS: A comma separated list of tags to add to all metrics. Blank by default.
  • COUNTDOWN_KEY: The name of the countdown metric. Defaults to countdown.
  • COUNTDOWN_ID: The name of the countdown metric tag. Defaults to name.
  • HEARTBEAT_METRIC: The name of the heartbeat metric. Defaults to heartbeat.

The metrics to watch are read as environment variables that start with the value of the ENV_PREFIX environment variable. The value of the environment variable should be a date in the format YYYY-MM-DDTHH:MM:SSZ. The prefix is stripped and the remaining text is used as the name of the metric.

For example, consider the following environment variables:

ENV_PREFIX=SURGE
METRIC_PREFIX=wildmagic_rocks
SURGE_FOO="2024-01-01T14:00:00-04:00"
WATCH_THIS_BAR="2024-01-01T14:00:00-04:00"

The above environment variable would result in a metric named wildmagic_rocks.foo with a value of the number of seconds until the date 2024-01-01T14:00:00-04:00 and a tag named name with a value of foo published every 10 seconds. The WATCH_THIS_BAR environment variable would be ignored because it is not one of the configuration environment variables nor does it have the prefix SURGE as defined by the ENV_PREFIX environment variable.

Example

First start a fake statsd sink:

$ nc -u -l -p 8125

Then run the daemon with an example metric:

$ COUNTDOWNS_FOO="2024-01-01T14:00:00-04:00" cargo run

You'll see metrics published with the default values:

countdown:22192887|g|#name:foo
heatbeat:1681939112|g
countdown:22192877|g|#name:foo
heatbeat:1681939122|g
countdown:22192867|g|#name:foo
heatbeat:1681939132|g

Global Tags

When running:

$ METRIC_TAGS="DD_SERVICE=cntdwn,DD_VERSION=1.2.0" COUNTDOWNS_FOO="2024-01-01T14:00:00-04:00" cargo run

You'll get:

countdown:22132947|g|#DD_SERVICE:cntdwn,DD_VERSION:1.2.0,name:foo
heatbeat:1681999052|g|#DD_SERVICE:cntdwn,DD_VERSION:1.2.0

Usage

Use this to monitor your stuff.

Deploy it with something like this:

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: cntdwn
  labels:
    app: cntdwn
spec:
  revisionHistoryLimit: 1
  selector:
    matchLabels:
      app: cntdwn
  template:
    metadata:
      labels:
        app: cntdwn
    spec:
      containers:
        - name: app
          image: ngerakines/countdown-metrics:v1.2.0
          env:
            - name: COUNTDOWNS_SITE_CERT
              value: "2024-01-01T14:00:00-04:00"
            - name: COUNTDOWNS_TS_KEY_DC1
              value: "2024-02-01T14:00:00-04:00"

Then create a monitor in your observability stack that alerts when the value of the metric is less than 345600 (4 days) or 950400 (11 days).

Dependencies

~10–20MB
~239K SLoC