#cron-job #runner #cron #scheduling #scheduled #schedule #run

cron_runner

Cron runner that runs your scheduled jobs

4 releases

0.1.3 Aug 12, 2024
0.1.2 Aug 12, 2024
0.1.1 Aug 12, 2024
0.1.0 Aug 12, 2024

#3 in #scheduled

MIT license

22KB
382 lines

cron_runner

Use the Job trait to define Jobs on your structs, pass the jobs to the Runner and start it with run(). The Runner will spawn a main "monitoring" thread to schedule your jobs, and additional worker threads to run your jobs.

The Runner can run in 2 modes.

Pooled

Spawn it with n > 0 threads and those threads will live in a thread pool waiting for new jobs from a mpsc channel ( in this context used like a FIFO queue). While the worker threads are waiting, they are blocked on recv(). If all workers are busy new jobs will not be run untill a worker is free, the first job to be scheduled is the first to be picked up and run by a worker.

Non Pooled

If you start the runner with 0 worker threads, any time a job should run a new thread will be spawned for the job that will execute it and return. If for some reason a worker could not be spawned, the Runner will try to run the job in its main thread, blocking scheduling for the rest of the jobs untill it is done.

Example

 use cron_runner::{Job, Runner, Schedule};

 struct PrintJob {
     message: String,
 }

 impl Job for PrintJob {
     fn execute(&self) {
         println!("{}", self.message);
     }

     fn schedule(&self) -> Schedule {
         "0 * * * * *".parse().unwrap()
     }
 }

 fn main() {
     let mut runner = Runner::new(2);

     runner.add_job(Box::new(PrintJob {
         message: "Print job".to_string(),
     }));

     runner.run();

     loop {
         std::thread::sleep(std::time::Duration::from_secs(5));
     }
 }

##Output

Output:
Worker started job 0 at 15:12:59.632829567
Print job
Worker finished job 0 at 15:12:59.632854926

Dependencies

~2–7MB
~48K SLoC