Suggest using CronJobs in Cron scaler docs for recurring scaling

[tomkerkhove]

Suggest using CronJobs in Cron scaler docs for recurring scaling which we don't support, but sometimes confuses people.

This is a scheduling need, not a scaling need.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 7 days if no further activity occurs. Thank you for your contributions.

[kai-66degrees]

@tomkerkhove I am working on a similar use case and I was exploring if KEDA fit the needs. What is the meaning of "recurring scaling which we don't support"? Can I not create a ScaledObject with a cron trigger for a deployment (like in the docs example) and scale it to the desired replicas/min replicas, and define another ScaledObject with a different timeslot and minReplicaCount to scale it again?

[tomkerkhove]

This is how our current CRON scheduler works:

See https://keda.sh/docs/latest/scalers/cron/#how-does-it-work.

This issue, however, is to document the scenario where you want to scale every x minutes/hours/etc.

[teebu]

I dont understand what this means. This statement is confusing.

What the CRON scaler does not do, is scale your workloads based on a recurring schedule.

Isn't a cron a recurring schedule? Meaning during the day I want my min pods ot scale to 10?

Can you clarify with a scenario on the statement what it doesn't mean. I think you're trying to say the end time is required for this to function.

[tomkerkhove]

Cron is used as a time notation when to start/end. By doing this, you define a window to which we will scale the workload correspondingly.

It does not mean, "scale every x min/h/days". For that, Kubernetes has CronJobs.