ceph (with examples)

1: Check cluster health status

ceph status

Motivation: Checking the cluster health status helps administrators identify any potential issues or problems with the Ceph storage system. It provides a quick overview of the overall health and functionality of the cluster.

Explanation: The ceph status command is used to display the current status of the Ceph cluster. It provides information about the number of OSDs (Object Storage Daemons), the number of Placement Groups (PGs), the status of the cluster services, and any warning or error messages.

Example Output:

  cluster:
    id:     12345678-90ab-cdef-1234-567890abcdef
    health: HEALTH_OK

  services:
    mon: 1 daemons, quorum a (age 1d)
    mgr: a(active, since 1d)
    osd: 3 osds: 3 up (since 1d), 3 in (since 1d)

  data:
    pools:   3 pools, 300 pgs
    objects: 1000 objects, 600 MiB
    usage:   3.0 GiB used, 57 GiB / 60 GiB avail
    pgs:     300 active+clean

2: Check cluster usage stats

ceph df

Motivation: Monitoring the cluster usage statistics helps administrators keep track of available storage, identify potential capacity issues, and plan for future expansion or optimization.

Explanation: The ceph df command provides detailed information about the disk space utilization in the Ceph cluster. It includes the total capacity, used space, available space, number of objects, and the utilization percentage for each storage pool.

Example Output:

RAW STORAGE:
    CLASS        SIZE     AVAIL      USED     RAW USED    % RAW USED
    ssd         1.0 TiB   400 GiB   600 GiB     900 GiB          90%
    hdd         5.0 TiB   2.4 TiB   2.6 TiB     5.0 TiB         100%

POOLS:
    NAME     ID     USED     %USED     OBJECTS
    pool1    1e     150 GiB    25%           500
    pool2    2d     400 GiB    40%          1200
    pool3    3f     300 GiB    30%           800
    pool4    4a     100 GiB    10%           400

3: Get statistics for placement groups in a cluster

ceph pg dump --format plain

Motivation: Understanding the distribution and state of placement groups (PGs) within the cluster can help administrators assess the load balancing, identify uneven distribution, and detect potential performance bottlenecks.

Explanation: The ceph pg dump --format plain command retrieves a detailed representation of the placement groups in the Ceph cluster. It provides information about the PG ID, state, acting and up OSDs, primary OSD, and the number of objects within each PG.

Example Output:

PG_STAT  OBJECTS       DEGRADED       MISPLACED            UNFOUND                BYTES        LOG DISK      LOG OSDS       STATES
    1               0                 0                     0                         0             0                    0           active+clean
    2              10                0                     0                     10 KB       12 KB              3             active+clean
    3           1000               10                 100                50 MB       150 MB              3            active+clean
    4         10000               10                  50                   100 GB      600 GB              3           active+clean

4: Create a storage pool

ceph osd pool create pool_name pg_num

Motivation: Creating storage pools allows administrators to organize and manage data in a Ceph cluster. They can allocate specific resources and set different storage policies for different applications, clients, or data types.

Explanation: The ceph osd pool create command creates a new storage pool in the Ceph cluster with the specified name and number of Placement Groups (PGs). The pool_name argument specifies the desired name for the pool, and the pg_num argument defines the number of PGs to create for the pool.

Example Code:

ceph osd pool create mypool 128

5: Delete a storage pool

ceph osd pool delete pool_name

Motivation: Deleting a storage pool helps administrators free up resources, reclaim disk space, or remove unnecessary data from the cluster. It is useful when a pool is no longer needed or when it needs to be reconfigured.

Explanation: The ceph osd pool delete command permanently removes a storage pool from the Ceph cluster. The pool_name argument specifies the name of the pool to be deleted.

Example Code:

ceph osd pool delete mypool

6: Rename a storage pool

ceph osd pool rename current_name new_name

Motivation: Renaming a storage pool allows administrators to update pool names to better reflect their purpose or to align with the changing needs of the environment. It simplifies management and ensures consistency within the cluster.

Explanation: The ceph osd pool rename command changes the name of an existing storage pool in the Ceph cluster. The current_name argument specifies the current name of the pool, and the new_name argument defines the desired new name for the pool.

Example Code:

ceph osd pool rename oldpool newpool

7: Self-repair pool storage

ceph pg repair pool_name

Motivation: Self-repairing pool storage helps to maintain data integrity and reliability in the Ceph cluster. It identifies and resolves any inconsistencies or errors within the placement groups, ensuring that data is accessible and intact.

Explanation: The ceph pg repair command initiates a self-repair process for a specific storage pool in the Ceph cluster. The pool_name argument specifies the name of the pool that needs repair. This command can be used when there are reported errors or inconsistencies within the pool.

Example Code:

ceph pg repair mypool

Conclusion

In this article, we explored different use cases of the ceph command. We covered how to check the cluster health status, monitor cluster usage statistics, retrieve placement group statistics, create, delete, and rename storage pools, as well as initiate self-repair for pool storage. These examples demonstrate the versatility and functionality of the ceph command in managing and maintaining a Ceph storage cluster.