mongo-keepalive (Rust)

Prevent MongoDB Atlas free-tier clusters from becoming inactive by periodically sending a ping command.

MongoDB Atlas pauses free-tier (M0) clusters after 60 days of inactivity. This library keeps your cluster alive by running db.adminCommand({ ping: 1 }) on a configurable interval (default: every 12 hours).

Installation

Add to your Cargo.toml:

[dependencies]
mongo-keepalive = "1.0.0"
tokio = { version = "1", features = ["rt-multi-thread", "macros"] }

Usage

use mongo_keepalive::start_keep_alive;

#[tokio::main]
async fn main() {
    let handle = start_keep_alive(
        "mongodb+srv://user:pass@cluster.mongodb.net/db",
        "12h",
    ).await.unwrap();

    // Keep running...
    tokio::signal::ctrl_c().await.unwrap();
    
    // Graceful shutdown
    handle.stop().await;
}

Features

⚡ Async/await - Built on Tokio for efficient concurrency
🔄 Automatic retry - 3 attempts with 5-second delays
📝 Logging - Uses the log crate for observability
🛡️ Error handling - Comprehensive error types
♻️ Graceful shutdown - Clean stop mechanism via handle

How It Works

Connects to MongoDB using the official async driver
Spawns a background task that pings every interval
Retries failed pings automatically
Logs all operations for monitoring

Configuration

Parameter	Type	Default	Description
`uri`	`&str`	-	MongoDB connection string (required)
`interval`	`&str`	`"12h"`	Ping interval (e.g. `"6h"`, `"30m"`)

Interval Format

"30m" → 30 minutes
"6h" → 6 hours
"12h" → 12 hours (default)

License

This project is licensed under the MIT License.

mongo-keepalive 1.0.1