Storage allocation
Introduction
In 22.1 the system setting storage_allocation
was introduced. This algorithm determines the way storage pools are selected when files are ingested or by the scheduler when exporting files to tape.
Types
FIRST_SYMMETRICAL
Within the primary cluster, we select the storage pool first storage pool with available space, as defined by the id in cluster
property. This typically corresponds with the natural order in which the pools where added. The primary cluster is typically a cluster of role ARCHIVE
. See Pools for additional information.
For each secondary cluster, we select the storage pool with the same ‘id in cluster’ property as the primary pool.
MOST_FILLED_SYMMETRICAL
Within the primary cluster, of the available storage pools with enough space, we select the pool with the least available space. Basically, this means we will fill one storage pool as much as possible before we start writing to new storage pools.
For each secondary cluster, we select the storage pool with the same ‘id in cluster’ property as the primary pool.
MOST_FILLED_ASYMMETRICAL
Within the primary cluster, of the available storage pools with enough space, we select the pool with the least available space. Basically, this means we will fill one storage pool as much as possible before we start writing to new storage pools.
For each secondary cluster, we will again select the pool with the least available space.