@kmcd03
"This cluster is scheduled to be decommissioned and VMs migrated to new VCF cluster in next 90 days. Is there any harm in ignoring the warning?"
Nope, no harm - the only thing not doing this will mean is objects will remain using older format and won't be able to avail of the new layout feature which permits using way less slack space while doing deep-reconfiguration of objects (e.g. when doing things like changing from Stripe-Width=1 to Stripe-Width=10).
"Or does anyone know if there are safeguards to prevent the change object format task from using all the free space?"
Yes, there are safeguards in place, it will pause resync of data to any disk that reaches 95% full (and at that point CLOM would be aggressively trying to move data off such a disk to a lower used disk), the relayout task can also of course just be cancelled.
"Will this task check for enough slack space to run?"
Well, not total but it does calculation before starting resync of each object and if it can't fit it then it won't start that object.
"Is it smart enough that it will only convert a few objects at a time and queue up the other objects?"
Yes, it does them few at a time - the only time this ever really has issues is in situations where there are very few nodes and disproportionally large objects (e.g. a 2-node cluster with 10TB per node and a single object using 6TB per replica).
