Currently for storage allocators, only random or firstfit or user-dispersing strategies are present. This is governed by the 'vm.allocation.algorithm' global config. So for you case with storage tags, if you choose random, either of the pools can get chosen. For firstfit, you will mostly see one pool getting filled up first and then the other.
There is no round-robin implementation in place. 'User-dispersing' may provide it to some effect if you are using a single account, since it always chooses a pool having less number of volumes for a given account. -Prachi -----Original Message----- From: Mike Tutkowski [mailto:mike.tutkow...@solidfire.com] Sent: Wednesday, July 10, 2013 1:34 PM To: dev@cloudstack.apache.org Subject: Storage Allocator Question Hi, I have a question about how current storage allocators work. Let's say I have two primary storages with the same storage tag. If I execute compute and disk offerings that reference that storage tag only, will one primary storage be (essentially) filled up before the other is utilized or do we have an allocator that performs round-robin placement? Thanks! -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloud<http://solidfire.com/solution/overview/?video=play> *(tm)*
