in⧕idents

Inxidents is a minimal configuration, open-source incident monitoring software with alerts and dashboard for your HTTP/S services written in Go.

Current Features:

• Real-time (SSE) Health Dashboard of your services. Perfect for office screens or similar environments.
• Slack Alerts whenever a service goes down 🟥 and recovers 🟩.
• ACK feature (acknowledge a down service will stop further notifications and will display corresponding service with black and yellow pattern 🚧)

• Types of checks: GET/POST, StatusCode, containsString (check if certain text is in the response body)
• Visually see the frequency of the healthcheck (the white progressbar animation)
• Small project with simple configuration. Easy to hack, deploy and further extend for your needs.

Upcoming features:

• Private/unique URLs for dashboards
• ... ideas and suggestions are welcome

Demo

Click for Demo Dashboard

Installation / Deployment

1. cp config.dev.yaml config.yaml
2. Change config.yaml accordingly and add your services: Example configuration of one service:

- name: Google
  endpoint: https://www.google.com
  frequency: 1m
  expectedCode: 200

• name: Name of service, currently it needs to be unique for each service you check.
• endpoint: HTTP/S endpoint
• frequency: Frequency of the health check, examples: "300ms", "2h45m". Valid time units are "ns", "us" (or "µs"), "ms", "s", "m", "h".
• expectedCode: This is the expected http status code returned from the endpoint.
• httpMethod: OPTIONAL - write POST if you are testing POST http Methods.
• containsString: OPTIONAL - Check if given string exists in the response body. Value type, string: "FAQ"
• disableAlerts: OPTIONAL - For some services one might want only the dashboard and not alerts, set true to those, default is false.
• userAgent: OPTIONAL - Set custom user-agent to requests checking the services.

3. To get Slack alerts, add an environmental variable called SLACK_WEBHOOK_URL containing the incoming slack webhook url. More info on it here

Deploy on fly.io

1. Install flytcl
2. Run flyctl launch(answer no to DB or Volume creations)
3. Run flyctl deploy to deploy

To enable Slack alerts when deploying to fly.io you can add the SLACK_WEBHOOK_URL in the fly.toml file

[env]
   SLACK_WEBHOOK_URL = "YOUR INCOMING SLACK WEBHOOK URL"

Deploy using Docker

Pull inxidents image from dockerhub:

docker pull piqoni/inxidents

Create a directory anywhere in you system and then put your inxidents config.yaml file, for example MYDIR/config.yaml.

Run the container (-e SLACK_WEBHOOK_URL is optional, only if you want alerts):

docker run \
  -p 8080:8080 \
  -v /PATH/TO/YOUR/MYDIR:/app \
  -e SLACK_WEBHOOK_URL=YOUR_SLACK_WEBHOOK_URL_HERE \
  piqoni/inxidents

Access the dashboard on http://localhost:8080

Tech comments / Architecture

There is no database by design for the time being (if needed in the future, it will likely be SQLite). Apart from the configuration file everything else happens in-memory. The only persistent data history (downtimes history) can be found on Slack alerts and application log files.

flowchart TB
  subgraph MainThread
    Main[read services in config.yaml]
  end

  subgraph Always Running Goroutines
    Service1[Service 1 Check]
    Service2[Service 2 Check]
    Service3[Service 3 Check]
    SendAlerts[When check fails/recovers]
  end

  Main -->|goroutine 1| Service1
  Main -->|goroutine 2| Service2
  Main -->|goroutine 3| Service3
  SendAlerts -->|Alert Message| Slack
subgraph Browser Dashboard
  Service1 -->|SSE Stream| EventSource
  Service2 -->|SSE Stream| EventSource
  Service3 -->|SSE Stream | EventSource
end