Setup AI Powered Metrics Services¶
Overview¶
This guide describes how to set up the following:
- Centralized metrics service with Prometheus and how to stream metrics to CloudAEye from environments such as Docker/Kubernetes
- View metrics in Grafana dashboard
- AI powered metrics analyzer service for anomaly detection
At the end of this tutorial, you will be able to setup metrics services for your distributed cloud applications.
Prerequisites¶
Before starting this tutorial, please install and configure caeops
if you plan to use CLI.
caeops
is a command line tool for working with CloudAEye SaaS (Software-as-a-Service).
Install and setup CloudAEye CLI.
:fontawesome-solid-sign: Register¶
Please sign up for CloudAEye SaaS.
Sign up with CloudAEye SaaS and request free trial.
Setup Metrics Service¶
After you get access to CloudAEye SaaS, do the following steps:
Step 1: Create Metrics Service¶
Create metrics service that will setup the server-side infrastructure for streaming metrics, and persistence of time series data.
Step 2: Setup Metrics Streaming¶
Using the CLI, setup metrics streaming from your microsrevices environment.
Step 3: View Metrics with Grafana¶
Setup Grafana service and view metrics in Grafna dashboard.
Setup Metrics Analyzer Service¶
Step 4: Setup AI Powered Metrics Analyzer¶
Before you begin, please run the metrics service for few days (7) so that there is enough metrics for the AI/ML models to train
Afer running the metrics service for few days, setup the metrics analyzer service. This will automatically identify anomaly in your cloud application and pin point KPIs that are critical for the production incident.