I mean it really depends on what the oncall looks like. If your system breaks every 5 minutes or you have other teams escalating non-urgent tasks constantly the

I mean it really depends on what the oncall looks like. If your system breaks every 5 minutes or you have other teams escalating non-urgent tasks constantly then 1 week on call seems crazy. But there are many companies (mine included) where on-call basically means: If services go down you have to figure out why. No bug fixing etc. Oncall purely from a infrastructure perspective.