Prometheus architecture
Prometheus server
- Retrieval - Pulls metrics data from target services like applications, servers and services ...etc
- Storage - Stores scraped metrics in a Time series DB
- HTTP server - web server displays data in an UI (prometheus web ui) which accepts PromQL query language
Levels of Monitoring
- Infrastructure level - CPU, RAM, Network ... etc
- Platform level - Kubernetes components, Docker, Bare server
- Application level - 3rd party application and Our own applications
What does it monitor ?
- CPU status
- Memory/Disk usage
- API requests count
- Exception count
Type of metrics
- Counter - How many times this event happened ? (ex: no. of exception, total no. of request a service as received)
- Gauge - What is the current value of this event now ? Values can go up and down (ex: current usage of RAM or DISK)
- Histogram - It is used to track how long an event took ? (eg: how long an API request took to complete)
Deploying types
- Manually configuring all the yml files (Not recommended)
- Using an operator for more info refer k8s operator
- Using an helm chart to deploy operator (Highly recommended approach)
helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update
helm install prometheus prometheus-community/kube-prometheus-stack