We can't disable those modules during the upgrade as it would lead to imbalanced clusters during an upgrade. How do we need to change those modules in order to cope with OSDs not being at the same version?
Pawan, do you have additional information, why this is required?
ok thank you Pawan. No, turn out this does in fact not apply to cephadm.
I am reopening this I think this one is still valid for Cephadm!
We want to avoid as much as PG movement during the upgrade and on the same lines we have this RFE - https://bugzilla.redhat.com/show_bug.cgi?id=2111224