ceph

distributed storage cluster management

TLDR

Show the overall cluster health

$ ceph status

Explain why the cluster is not healthy

$ ceph health detail

Watch cluster events as they happen

$ ceph -w

Show capacity, overall and per pool

$ ceph df

List the pools

$ ceph osd pool ls

Create a pool with a placement group count

$ ceph osd pool create [pool_name] [128]

Change a pool's replica count

$ ceph osd pool set [pool_name] size [3]

Rename a pool

$ ceph osd pool rename [current_name] [new_name]

Delete a pool, which insists on the name twice and a confirmation flag

$ ceph osd pool delete [pool_name] [pool_name] --yes-i-really-really-mean-it

Show the OSDs arranged by host and rack

$ ceph osd tree

Mark an OSD out so its data drains onto the others

$ ceph osd out [osd.3]

Summarize the placement group states

$ ceph pg stat

Repair a specific placement group

$ ceph pg repair [pg_id]

Show the monitor quorum

$ ceph mon stat

ceph is the command-line interface for managing Ceph distributed storage clusters. Ceph provides unified object, block, and file storage with automatic data replication, self-healing, and no single point of failure.The command communicates with the Ceph monitor daemons to perform cluster administration tasks including checking health status, managing OSDs (Object Storage Daemons), configuring placement groups, and handling authentication. It serves as the primary tool for day-to-day cluster operations and troubleshooting.Ceph organizes data into pools, which contain placement groups that map to OSDs across the cluster. The `ceph status` command provides a quick overview of cluster health, while more specific subcommands allow fine-grained control over individual components.

PARAMETERS

-s, --status

Show cluster status

--format format

Output format: json, plain

-c file

Use alternate config file

CONFIGURATION

/etc/ceph/ceph.conf

Main cluster configuration including monitor addresses, authentication settings, and OSD parameters.

/etc/ceph/ceph.client.admin.keyring

Admin authentication keyring for cluster access.

SUBCOMMANDS

status

Show cluster health and status

Show cluster disk usage

osd

Manage Object Storage Daemons

Manage placement groups

mon

Manage monitor daemons

mds

Manage metadata servers

auth

Manage authentication

INSTALL

sudo apt install ceph-common

sudo dnf install ceph-common

sudo apk add ceph18-common

sudo zypper install ceph-common

nix profile install nixpkgs#ceph

CAVEATS

Requires proper Ceph configuration and authentication. Cluster operations may take time to complete. Some commands require specific permissions.