Pod Resources Monitoring

This project is a Kubernetes resource monitoring application that retrieves pod metrics and checks resource usage periodically. It sends alerts and warnings based on the defined resource limits and requests for each container within the pods.

Features

Monitors CPU and memory usage of Kubernetes pods.
Sends alerts if resource usage exceeds defined limits or requests.
Supports Slack notifications for alerts.
Configurable monitoring duration.

Requirements

Go 1.16 or later
Kubernetes cluster
Access to Kubernetes API
Slack account (optional for notifications)

Installation

Clone the repository:

git clone https://github.com/saidsef/pod-resources.git
cd pod-resources

Install the required Go modules:

go mod tidy

Configuration

Before running the application, you need to set up the following environment variables - all optional:

DURATION_SECONDS: The duration (in seconds) for which the application will check resource usage. Default is 120s.
SLACK_TOKEN: The token for your Slack app to send notifications (optional).
SLACK_CHANNEL: The Slack channel where notifications will be sent (optional).

Usage

To run the application, execute the following command:

go run resources.go

The application will connect to the Kubernetes cluster, retrieve the list of pods, and start monitoring their resource usage based on the specified duration.

Alerts and Notifications

If a container exceeds its resource request, an alert will be sent.
If a container has a limit set but no request defined, a warning will be logged.
If a container exceeds its resource limit, an alert will be sent.
Warnings will be logged if no limits or requests are defined for CPU or memory.