Hi Wissem
Thank you for your reply.
As the erasure-code-profile is a pool setting, it is on a lower layer and rgw should be
unaware and independent of it, but it could play role regarding this know issue with space
allocation and EC pools.
Anyway this does so seem to be the cause here; after running rgw-orphan-list it appears
that the secondary had about 80 M orphan objects.
I do not know the reason yet but I will restart the sync from scratch and investigate
further with the logs...
(The other Issue with the wrong size_utilized was there long before we created the
secondary zone)
Cheers
Francois
--