Rate This Document
Findability
Accuracy
Completeness
Readability

Environment

Physical Networking

The physical environment of Ceph devices adopts a scheme of "two network layers + three nodes." MON, MGR, MDS, and OSD nodes are deployed together. The public network is separated from the cluster network. The two networks use 25GE optical ports for communication.

The Ceph cluster consists of Ceph clients and Ceph servers. Figure 1 shows the physical networking. Configure the networking based on site requirements.

Figure 1 Physical networking

Hardware Configuration

Table 1 shows the Ceph hardware configuration.

Table 1 Hardware configuration

Server

Kunpeng server

Processor

Kunpeng 920 processor

Core

2 x 32 cores

CPU Frequency

2600 MHz

Memory Capacity

12 x 16 GB

Memory Frequency

2666 MHz (8 x Micron 2R memory module)

NIC

IN200 NIC (4 x 25GE)

Drive

System drives: RAID 1 (2 x 960 GB SATA SSD)

Data drives of general-purpose storage: JBOD enabled in RAID mode (12 x 4 TB SATA HDD)

NVMe SSD

Acceleration drive of general-purpose storage: 1 x 3.2 TB ES3600P V5

Data drives of high-performance storage: 12 x 3.2 TB ES3600P V5

RAID Controller Card

Avago SAS 3508

Software Versions

Table 2 lists the required software versions.

Table 2 Software versions

Software

Version

OS

CentOS Linux release 7.6.1810

openEuler 20.03 LTS SP1

openEuler 22.03 LTS SP2

Ceph

14.2.8 or 14.2.10 is recommended.

ceph-deploy

2.0.1

CosBench

0.4.2.c4

Node Information

Table 3 describes the IP network segment planning of the hosts.

Table 3 Node information

Host Type

Host Name

Public Network Segment

Cluster Network Segment

OSD/MON node

node1

192.168.3.0/24

192.168.4.0/24

OSD/MGR node

node2

192.168.3.0/24

192.168.4.0/24

OSD/MDS node

node3

192.168.3.0/24

192.168.4.0/24

Component Deployment

Table 4 describes the deployment of service components in the Ceph block device cluster.

Table 4 Component deployment

Physical Machine Name

OSD

MON

MGR

node1

12 OSDs

1 MON

1 MGR

node2

12 OSDs

1 MON

1 MGR

node3

12 OSDs

1 MON

1 MGR

Cluster Check

Run the ceph health command to check the cluster health status. If HEALTH_OK is displayed, the cluster is running properly.