nf ceph fs
The nf ceph fs command creates a new CephFS filesystem. CephFS requires at least one MDS daemon to be running (see nf ceph mds).
Each CephFS requires two pools: a metadata pool for directory and inode information, and a data pool for file contents. You can control the placement group count for each pool — lower PG counts work for small clusters, while larger clusters benefit from more PGs.
Usage
$ nf ceph fs --name <fs-name> --metadataPool <pool> --dataPool <pool>
Options
nf ceph fs
nf ceph fsCreate a new CephFS filesystem with dedicated metadata and data pools.
| Flag | Description |
|---|---|
--name | Name of the CephFS (required) |
--metadataPool | Metadata pool name (required) |
--metadataPg | Metadata pool placement groups (default: 16) |
--dataPool | Data pool name (required) |
--dataPg | Data pool placement groups (default: 32) |
Terminals
$ nf ceph fs --name myfs --metadataPool myfs_meta --dataPool myfs_data CephFS creation triggered.
$ nf ceph fs --name production --metadataPool prod_meta --metadataPg 64 --dataPool prod_data --dataPg 128 CephFS creation triggered.