Description of problem:
Disallow enabling 'hashpspool' option to a pool without some kind of --i-understand-this-will-remap-all-pgs flag
Upstream Feature Tracker: http://tracker.ceph.com/issues/18468
Some customers have clusters which were created during bobtail release
and at that time this option 'hashpspool' was not present.
We need this flag ' --i-understand-this-will-remap-all-pgs ' because if any user enables 'hashpspool' option on a pool, it will change the way in which PGs in that pool are hashed(it will basically remap every pg in the pool)
and then the data movement would be catastrophic.
Other than pg hashing for the way pool/pg overlap works there are no other benefits
and for a production cluster which has a large number of osds and a large amount of data
enabling this option is surely not recommended at all.
Version-Release number of selected component (if applicable):
Red Hat Ceph Storage 1.3.3
Master PR: https://github.com/ceph/ceph/pull/13406
Backport PR for 10.2.10: