Feature #5134

OpenNebula should not assume total available data in Ceph cluster is available for ONE

Added by Stefan Kooman almost 4 years ago. Updated over 3 years ago.

Status:PendingStart date:04/27/2017
Priority:NormalDue date:
Assignee:-% Done:


Target version:-
Resolution: Pull request:


OpenNebula Ceph monitor assumes all data in the Ceph cluster is available for ONE, i.e. max_avail. Monitor should check if there is a quota assigned to the pool and if so, use this value instead of max_avail. See https://github.com/OpenNebula/one/pull/127 for a possible fix.


#1 Updated by Ruben S. Montero almost 4 years ago

You can set LIMIT_MB, to limit maximum capacity allowed for the datastore. Could that address this issue?

#2 Updated by Stefan Kooman over 3 years ago

It's more about monitoring than it's about not able to limit usage. A cloudadmin that is checking the datastores wants to see the amount of available storage, not some theoretical maximum. When datastore attributes "TOTAL_MB", "FREE_MB", "USED_MB" are used for monitoring a cluster, you also want to get the right values.

Also available in: Atom PDF