Minimum hardware considerations

Before running IBM Storage Ceph, be sure to comply with all minimum hardware requirements listed here.

IBM Storage Ceph can run on non-proprietary commodity hardware. Small production clusters and development clusters can run without performance optimization with modest hardware.
Note: Disk space requirements are based on the Ceph daemons' default path under /var/lib/ceph/ directory.
For more information about the IBM Storage Ceph internal components and the strategies around those components, see Storage strategies.
Important: Hardware accelerated compression in Ceph Object Gateway requires RHEL 9.4 on a Sapphire or Emerald Rapids Xeon CPU (or newer) with QAT devices. For more information, see Which 4th Gen Intel® Xeon® Scalable Processors Do Support Intel® QuickAssist Technology (Intel® QAT)? on Intel Product Support.
IBM Storage Ceph is supported on Intel and AMD x86-64 microprocessors.
  • IBM Storage Ceph 9.9.0 introduces full support for clusters built with ARM (aarch64).

Use the following information for each process container and their minimum recommended requirements for various components.

For the latest architectural requirements for a supportable cluster, see IBM Storage Ceph Supported configurations .

ceph-osd-container
  • 1 × AMD64 or Intel 64 CPU vcore/hyperthread per OSD container.
  • Minimum of 5 GB of RAM per OSD container.
  • 1 × OS disk per host.
  • 1 × storage drive per OSD container.
    Note: The storage drive cannot be shared with the OS disk.
  • Optional WAL and DB offload for HDD OSDs.
    • IBM recommended.
    • 1 × SSD or NVMe or Optane partition or lvm per daemon.
    • Sizing is 4% of block.data for BlueStore for object, file, and mixed workloads and 1% of block.data for the BlueStore for Block Device.
  • Minimum of 2 × 10 GB Ethernet NICs network connection.
ceph-mon-container
  • 1 × AMD64 or Intel 64 CPU vcore/hyperthread per mon-container.
  • 3 GB of RAM per mon-container.
  • 10 GB of disk space per mon-container.
    Note: 50 GB disk space is recommended.
  • Prometheus requires 20 GB to 50 GB under /var/lib/ceph/ directory created as a separate file system to protect the contents under /var/ directory.
ceph-mgr-container
  • 1 × AMD64 or Intel 64 CPU vcore/hyperthread per mgr-container.
  • 3 GB RAM per mgr-container.
  • 2 × 1 GB Ethernet NICs network connection.
    Note: 10 GB network is recommended.
ceph-radosgw-container
  • 1 × AMD64 or Intel 64 CPU vcore/hyperthread per radosgw-container.
  • 1 GB RAM per daemon.
  • 5 GB disk space per daemon.
  • 1 × 1 GB Ethernet NICs network connection.
ceph-mds-container
  • 1 × AMD64 or Intel 64 CPU vcore/hyperthread per mds-container.

    This number is highly dependent on the configurable MDS cache size. The RAM requirement is typically twice as much as the amount set in the mds_cache_memory_limit configuration setting.

    Note: This is the memory for the daemon, not the overall system memory.
  • 3 GB RAM per mds-container.
  • 2 GB disk space per mds-container, plus considering any additional space required for possible debug logging.
    Note: Minimum of 20 GiB of available space in the CephFS metadata pool.
  • Minimum of 2 × 1 GB Ethernet NICs network connection.
    Note:
    • 10 GB network is recommended.
    • This is the same network as the OSD containers. If you have a 10 GB network on your OSDs you should use the same on your MDS so that the MDS is not disadvantaged when it comes to latency.
nvmeof
  • 4 × Intel 64 CPU vcore/hyperthread CPUs (6th generation or later with AWX512 support).
  • 8 GB RAM per gateway on a Ceph node.
Note: For NVMe-oF best practices, see NVMe-oF gateway performance best practices.