rlcheck - Rust live check tool

A Rust program that live checks web endpoints for uptime and content changes.

• Use this to make serverless endpoints always hot/warm.
• Designed to be simple and lightweight to be able to run even on SBCs.

Features

• Checks if web endpoints are up (HTTP status)
• Detects content changes using MD5 hashing
• Configurable check intervals per site
• YAML configuration file
• Monitors multiple sites concurrently
• Optional logging to file with automatic log rotation
• Log rotation: maintains up to 4 log files with 20K lines each

Configuration

Create a config.yaml file with the following format:

sites:
  - url: "https://example.com"
    interval: 30  # check every 30 seconds
  - url: "https://another-site.com"
    interval: 60  # check every 60 seconds

Config Fields

• url: The web enpoint URL to monitor (must include protocol: http:// or https://)
• interval: How often to check the site (in seconds)

Usage

make install
# installs rlcheck binary under: $HOME/.cargo/bin
# make sure $HOME/.cargo/bin is in your $PATH
cp config.yaml config.prod.yaml
# update config.prod.yaml file with url and interval
make prod

# Tested on Mac and Linux

Build the project

cargo build --release

Run with default config (config.yaml)

cargo run

Run with custom config file

cargo run -- --config path/to/your-config.yaml

Run with logging enabled

cargo run -- --log-file monitor.log

Command-line options

• -c, --config <FILE>: Path to config file (default: config.yaml)
• -l, --log-file <FILE>: Path to log file (optional)
• -h, --help: Print help information

Log Rotation

When logging is enabled, the monitor automatically rotates log files:

• Maximum 4 log files maintained (monitor.log, monitor.log.1, monitor.log.2, monitor.log.3)
• Each log file can contain up to 20,000 lines
• When the current log reaches 20K lines, it's rotated:
  • monitor.log → monitor.log.1
  • monitor.log.1 → monitor.log.2
  • monitor.log.2 → monitor.log.3
  • monitor.log.3 is deleted
  • New monitor.log is created

Output

The program displays simple, structured output on a single line with these key fields:

• url: The URL being monitored
• load_time: Response time in milliseconds
• status: up, down, or error
• size: Content size in bytes
• content_hash: First 5 characters of the MD5 hash
• error: Error message (only for error status)

Example Output

Reading config from: config.yaml

Monitoring 2 site(s):
  - https://example.com (check every 30s)
  - https://httpbin.org/status/200 (check every 60s)

Starting monitoring... (Press Ctrl+C to stop)

url: https://example.com | load_time: 143ms | status: up | size: 18585bytes | content_hash: 05f75

url: https://httpbin.org/status/200 | load_time: 305ms | status: up | size: 0bytes | content_hash: d41d8

url: https://example.com | load_time: 167ms | status: up | size: 18627bytes | content_hash: a3f8d

url: https://httpbin.org/status/200 | load_time: 298ms | status: up | size: 0bytes | content_hash: d41d8

url: https://example.com | load_time: n/a | status: error | error: Request failed: connection timeout

Dependencies

• reqwest - HTTP client
• serde & serde_yaml - Configuration parsing
• md5 - Content hashing
• tokio - Async runtime
• clap - Command-line argument parsing

Binary Optimization

The release binary is optimized for size:

• Size: ~1.7MB (stripped)
• Optimizations: LTO enabled, size optimization (opt-level = "z"), stripped symbols
• TLS: Uses rustls instead of native-tls for smaller footprint
• Minimal tokio runtime with only required features

Notes

• The program runs indefinitely until stopped with Ctrl+C
• Each site is monitored concurrently in its own async task
• Content changes are detected by comparing MD5 hashes of the response body
• All output is on a single line per check for easy parsing and logging
• Content size is reported in bytes for every request
• Errors include the error message directly in the output line

License

MIT License - see LICENSE file for details