What you will learn in this TP
- Install cAdvisor
- Access cAdvisor's console and metrics
- Install Node exporter
- Access Node exporter's console and metrics
Prerequisites #
You must have a machine with Docker installed. If you don't have a machine or test environment, please complete the 'setting up the test environment' lab first.
cAdvisor #
What you will learn in this section
- Install cAdvisor
- Access cAdvisor's console and metrics
▶ Review the concept presentation video
Install and configure cAdvisor
Deploy the container
You must deploy a cAdvisor container to collect all the container metrics present on the machine. It must be added to the observability network. Find an example docker-compose file with the configuration to deploy cAdvisor here: File.Check that it's working
Verify that the exporter is working correctly and responds on its listening port.
Check the metrics
Verify that the metrics are coming through correctly
Node exporter #
What you will learn in this section
- Install Node exporter
- Access Node exporter's console and metrics
▶ Review the concept presentation video
Deploy the container
You must deploy a Node_exporter container to collect all of the machine's metrics. It must be added to the observability network. Find an example docker-compose file with the configuration to deploy cAdvisor here: File.Check that it's working
Verify that the exporter is working correctly and responds on its listening port.
Check the metrics
Verify that the metrics are coming through correctly
Difficulty level : ●●●○○ (3/5)
My progress
Sign in to track your progress and take notes. Sign in
Course Glossary
Alerting
The process of notifying teams (via email, Slack, Teams) when a specific event, threshold breach, or anomaly is detected.
Architecture
The overall structural design of an IT system, defining its various components, their relationships, and the principles guiding its evolution.
OpenTelemetry
An open source framework providing standards, APIs, and SDKs to collect and export observability data (traces, metrics, logs) in an agnostic way.
Recommended Articles
Prometheus dans le monde Kubernetes (Operator, ServiceMonitor, kube-prometheus-stack)
Superviser Kubernetes avec Prometheus : service discovery, kube-prometheus-stack, Prometheus Oper...
Mimir : ruler, Alertmanager et mimirtool
Évaluer des règles d'enregistrement et d'alerte dans Mimir, configurer l'Alertmanager multi-tenan...
Alertmanager : rôle, architecture et cycle de vie d'une alerte
Ce que fait Alertmanager entre Prometheus et vos équipes : déduplication, regroupement, routage, ...
Écrire de bonnes règles d'alerte Prometheus (et les tester)
Concevoir des alertes utiles : alerter sur les symptômes, choisir for et les seuils, labels de sé...
Haute disponibilité et exploitation d'Alertmanager
Monter un cluster Alertmanager hautement disponible (gossip), configurer Prometheus pour tous les...
Bonnes pratiques d'alerting : Prometheus, Alertmanager et astreinte
Check-list pour un alerting efficace : règles sur les symptômes, sévérités, routage par équipe, r...
Mimir : le chemin d'écriture (distributor, ingesters, réplication et déduplication HA)
Comment Mimir ingère les métriques : remote_write, validation et limites du distributor, hash rin...
Thanos Query et Query Frontend : une vue globale et dédupliquée
Configurer Thanos Query : endpoints StoreAPI, déduplication par replica label, partial response, ...
Thanos Ruler : règles et alertes sur la vue globale
Évaluer des règles d'enregistrement et d'alerte sur l'ensemble des clusters avec Thanos Ruler : c...