Rackspace Cloud Monitoring agent plugin for Varnish
Place varnish.sh into /usr/lib/rackspace-monitoring-agent/plugins/ on each Varnish server.
Create a Cloud Monitoring Check of type "agent.plugin". Also specify the following values in the details argument:
| Check Attribute | Value | Description |
|---|---|---|
| file | varnish.sh | The file name of the plugin script |
| args | cache_hit,cache_hitpass,cache_miss | Comma-separated list of valid varnishstat metrics |
| Metric Name | Type | Description |
|---|---|---|
| hit_percent | float | 100 - (cache_miss/cache_hit) |
| healthy_backends | int32 | Number of backends reporting healthy from varnishadm debug.health |
| args (above) | gauge | Current value of each varnishstat metric passed in |
Create a Cloud Monitoring Alarm that alarms on:
if (metric['healthy_backends'] < 1) {
return new AlarmStatus(CRITICAL, 'Varnish doesnt have any backends!');
}
if (metric['healthy_backends'] < 2) {
return new AlarmStatus(WARNING, 'Varnish only has #{healthy_backends} healthy backend.');
}
return new AlarmStatus(OK, 'Varnish has \#{healthy_backends} backends.');