It’s not clear how to start distributed workers

[cristiano-belloni]

From the site, the impression is that Sidequest handles distributed workers running on different machines. Reading the documentation, it’s not clear how workers are spawned and where. The impression is that they run on the same machines where Sidequest is running. Is that accurate? If yes, how “Spin up as many workers as you need across multiple nodes” on your front page holds?

[GiovaniGuizzo]

Hey, thanks for the interest in our OSS :)

A worker is spawned anytime you call Sidequest.start in a process. So, if you want to spawn multiple workers, you simply call Sidequest.start on any machine you want to transform in a worker and make sure they all point to the same database.

That's about it. As long as the workers have access to the same scripts used to enqueue, they will be able to run the jobs.

Take another look at this.

[GiovaniGuizzo]

A side note, you can run X workers on a single machine if you call Sidequest.start on X processes.