In a FabricPool configuration, the tiered cold data will be moved to performance tier from cloud/capacity tier when accessed for volumes with tiering policies snapshot-only and auto. For volumes with 'all' tiering policy, all data will be served directly from cloud/capacity tier.
However, if the performance tier USED is 70% or above, then the data from cloud tier will not be moved to performace tier upon access for any tiering policy and the following EMS will be triggered:
Aggregate aggr2_ssd: Current available space in the performance tier is below 30%%. Data migration from the capacity tier to the performance tier is stopped.
As a result, all the data will be served directly from the capacity/cloud tier and you may experience latencies with IO access. Increase the available space on the aggregate so the data can be moved from capacity tier to performance tier when accessed.