Ceph Luminous Installation Guide

Environment Description

Role	Host Name	Ip	OS	Hard disk
Admin	Admin	10.0.0.230	CentOS 7.4
Mon & OSD & Mgr & MDS	node231	10.0.0.231	CentOS 7.4	/dev/vda/dev/vdb
Mon & OSD & Mgr	node232	10.0.0.232	CentOS 7.4	/dev/vda/dev/vdb
Mon & OSD & Mgr	node233	10.0.0.233	CentOS 7.4	/dev/vda/dev/vdb
Client	Client	10.0.0.234	CentOS 7.4

Set the host name of each server as shown in the table above.

Hostnamectl Set-hostname Host Name

The following operations need to be performed on all nodes.

Stop the firewall

Systemctl Disable FIREWALLD

Systemctl Stop Firewalld

SELinux is prohibited.

Vim/etc/selinux/config

Configured as Disabled

Configure/etc/hosts

10.0.0.230 Admin

10.0.0.231 node231

10.0.0.232 node232

10.0.0.233 node233

10.0.0.234 Client

Replace Yum source as domestic Alibaba cloud

http://blog.csdn.net/chenhaifeng2016/article/details/78864541

Adding a Ceph installation source

Vim/etc/yum.repos.d/ceph.repo

[Ceph]
Name=ceph packages for $basearch
baseurl=http://mirrors.aliyun.com/ceph/rpm-luminous/el7/$basearch
Enabled=1
gpgcheck=1
type=rpm-md
gpgkey=https://download.ceph.com/keys/release.asc

[ Ceph-noarch]
name=ceph noarch packages
Baseurl=http://mirrors.aliyun.com/ceph/rpm-luminous/el7/noarch
enabled=1
gpgcheck=1
type=rpm-md
gpgkey=https://download.ceph.com/keys/release.asc

[ Ceph-source]
name=ceph Source Packages
Baseurl=http://mirrors.aliyun.com/ceph/rpm-luminous/el7/srpms
enabled=1
gpgcheck=1
type=rpm-md
GPGKEY=HTTPS://DOWNLOAD.CEPH.COM/KEYS/RELEASE.ASC

Installing NTP

Yum Install NTP

Systemctl Enable NTPD

Systemctl Start NTPD

View NTP status

Ntpq-p

Restart all nodes

Shutdown-r Now or reboot

The following operations are only required to run on the admin node

Configure SSH password-free login

Ssh-keygen

Ssh-copy-id Admin

Ssh-copy-id node231

Ssh-copy-id node232

Ssh-copy-id node233

Ssh-copy-id Client

Installing Ceph-deploy

Yum Install Ceph-deploy

To create a configuration file directory

Mkdir-p/etc/ceph

Cd/etc/ceph

Create a Ceph cluster

Cepy-deploy New node231

Installing a Ceph binary package on all nodes

Ceph-deploy Admin node231 node232 node233 Client

Ceph-v or Ceph version

Create a Ceph MON

Ceph-deploy Mon create-initial

Create a Ceph OSD

Ceph-deploy Disk List node231

Ceph-deploy Disk Zap Node231:vdb

Ceph-deploy Disk Zap Node232:vdb

Ceph-deploy disk Zap Node233:vdb Ceph-deploy--overwrite-conf OSD Create Node231:vdb ceph-deploy--overwrite-conf OSD Crea Te node232:vdb ceph-deploy--overwrite-conf OSD Create Node233:vdb
Edit File/etc/ceph/ceph.conf Add the following public_network = 10.0.0.0/24

Copy the configuration file to each node Ceph-deploy admin client node231 node232 node233

Create Mon ceph-deploy--overwrite-conf Mon create node231 ceph-deploy--overwrite-conf admin node231
Ceph-deploy--overwrite-conf Mon Create node232 ceph-deploy--overwrite-conf admin node232
Ceph-deploy--overwrite-conf Mon Create node233 ceph-deploy--overwrite-conf admin node233
At this time, Mon and OSD were established on 3 nodes. Check the status


There are 3 Mon, 3 OSD, but the status of cluster is Health_warn, because it is no active MGR, next create Ceph Mgr
Ceph-deploy Mgr Create node231 ceph-deploy Mgr Create node232 ceph-deploy Mgr Create node233


In this step, the Ceph cluster has been installed.
Next Test the block storage, the following operations are performed on the client node.

Create a new storage pool instead of using the default RBD ceph OSD Pool Create test 128
Create a block RBD create--size 10G disk01--pool test
View RBD


View the properties of a block RBD info--pool Test disk01

Because the kernel is not supported, some features need to be suppressed, only layering is reserved

RBD--pool Test feature disable Disk01 exclusive-lock, Object-map, Fast-diff, Deep-flatten

Map Block disk01 to Local

RBD Map--pool Test disk01

Format Block devices

Mkfs.ext4/dev/rbd0

Mount the rbd0 to a local directory

Mount/dev/rbd0/mnt

This time to view the cluster status, the status of the cluster is Health_warn

Perform Ceph health Detail

Perform ceph OSD Pool application Enable test RBD as prompted

The cluster status is OK.