How do I quickly create an anomaly to see this framework alerts?

[kunapuli09]

I have telegraf, influxdb up and running along with morgoth with the following config. I don't see any anomolies even when the cpu spikes are 100%. I am seeing alertmanager reporting no anomalies. Is there a quick way to create an anomaly using your example docs.

schedules:

• query: SELECT value FROM "day".cpu_usage_idle GROUP BY * period: 30s delay: 10s tags: ret: day
• query: SELECT value FROM "week".cpu_usage_idle GROUP BY * period: 5m delay: 1m tags: ret: week
• query: SELECT value FROM "month".cpu_usage_idle GROUP BY * period: 1h delay: 1m tags: ret: month

mappings:

• name: cpu_.* tags: ret: month detector: fingerprints:
  • sigma: deviations: 10 alerts:
• query: SELECT COUNT(start) FROM anomaly GROUP BY * message: Too many anomalies detected threshold: 1 period: 1m group_by_interval: 2m notifiers:
  • log: file: alerts.log

[kunapuli09]

I am seeing some anomalies after running it for few hours. Each one of those recorded a bad date time stamps with epoch date (1970-01-01T00:00:00Z ) which is a different issue. I am closing this.