Looks really interesting, especially with the dashboards.
Just to validate an idea here: I’m using k8s (20+ services) and trying to stick to the 12factor design pattern by having all the baking/companion services, also like cron jobs deployed from within the service directory. Right now I’m using k8s cron services for cron jobs and log the steps and observe with DataDog. Using k8s cron services feels right somehow but the observability with DataDog not. So, would it make sense for me, to deploy a baking k8s web service running sidequest instead?
Well, Sidequest comes with a comprehensive built-in dashboard that provides real-time monitoring, job management, and performance analytics, so there's that.
You can probably deploy Sidequest a job processor for all your 20+ services. Each service can enqueue the jobs and you can run as many machines as you want with Sidequest just to run the jobs. You maintain the 12-factor principle - your services remain stateless and delegate scheduled work to Sidequest.
The only requirement here is that the scripts to be executed must be in the same path in all machines. As long as you follow that, you can deploy it and run your jobs :)