What happens when two aggregates are rejoined?

Rejoining aggregates is a supported Data ONTAP operation; the concern is not impossibility but rather the guaranteed data loss it causes.

Data ONTAP does not selectively preserve the newer aggregate's data - one aggregate is unconditionally overwritten regardless of relative age.

Aggregate rejoin does not roll both aggregates back to a common snapshot checkpoint; only one aggregate is overwritten to match the other's state at split time.

Rejoining aggregates does not produce a merged aggregate containing all volumes from both sides; one side's post-split changes are simply discarded.

During an aggregate rejoin, Data ONTAP reverts one aggregate back to the exact state it was in at the moment of the original split, overwriting all changes made to that aggregate afterward. This is a destructive, unrecoverable operation - any data written to the overwritten aggregate between the split and the rejoin is permanently gone.