Small blocks policy settings

This feature is available only when the special devices group exists in the pool.

Devices assigned to the special devices group are designated for storing specific data, including metadata, indirect blocks of user data, and deduplication tables. Additionally, devices in the special devices group can be configured to handle small file blocks that are not listed above by applying the small blocks policy.

The size of the small block refers to the size of a single block of data configured on the dataset. Maximum size of such blocks can be set for each dataset under the “Record size” option.

Deduplication tables can alternatively be placed in a separate group known as the deduplication group.

WARNING: If the size of the small block is greater than or equal to the value of record size on the dataset, all the blocks will be offloaded to the special devices group.

The small block size can be configured for the whole Pool or for each dataset separately. Options to choose range from 4 KiB to 16 MiB.

Configuring a bigger size for the small block policy can be helpful in case the administrator expects the to have a substantial amount of small files that will require low access times separated from the bigger files. Using this option is recommended only if the administrator understands what kind of data will be stored in configured datasets and the maximum size of offloaded files is exactly known, to avoid accidental data offload and special devices congestion.