192.168.0.210 deploy 192.168.0.211 node1 192.168.0.212 node2 192.168.0.213 node3
（1）关闭firewalled
systemctl stop firewalld
systemctl disable firewalld
（2）关闭selinux：
sed -i ‘s/enforcing/disabled/‘ /etc/selinux/config setenforce 0
（3）关闭NetworkManager
systemctl disable NetworkManager && systemctl stop NetworkManager
（4）添加主机名与IP对应关系：
vim /etc/hosts 172.30.112.78 deploy 172.30.112.179 node1 172.30.112.115 node2 172.30.112.82 node13
（5）设置主机名：
hostnamectl set-hostname deploy hostnamectl set-hostname node1 hostnamectl set-hostname node2 hostnamectl set-hostname node3
（6）同步网络时间和修改时区
yum install chrony -y systemctl restart chronyd.service && systemctl enable chronyd.service cp /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
（7）设置文件描述符
for i in deploy node1 node2 node3;do ssh $i "echo "ulimit -SHn 102400" >> /etc/rc.local";done cat >> /etc/security/limits.conf << EOF * soft nofile 65535 * hard nofile 65535 EOF
（8）内核参数优化
for i in deploy node1 node2 node3;do ssh $i "echo ‘vm.swappiness = 0‘ >> /etc/sysctl.conf";done for i in deploy node1 node2 node3;do ssh $i "echo ‘kernel.pid_max = 4194303‘ >> /etc/sysctl.conf";done for i in deploy node1 node2 node3;do ssh $i "sysctl -p";done
（9）在deploy上配置免密登录到node1 node2 node3
for host in node{1..3}; do ssh-copy-id $host;done
(10)read_ahead,通过数据预读并且记载到随机访问内存方式提高磁盘读操作
echo "8192" > /sys/block/sda/queue/read_ahead_kb
(11) I/O Scheduler，SSD要用noop，SATA/SAS使用deadline
echo "deadline" >/sys/block/sd[x]/queue/scheduler echo "noop" >/sys/block/sd[x]/queue/scheduler
vim /etc/yum.repos.d/ceph.repo [ceph] name=ceph baseurl=http://mirrors.aliyun.com/ceph/rpm-nautilus/el7/$basearch enabled=1 gpgcheck=1 type=rpm-md gpgkey=https://download.ceph.com/keys/release.asc [ceph-noarch] name=cephnoarch baseurl=http://mirrors.aliyun.com/ceph/rpm-nautilus/el7/noarch/ enabled=1 gpgcheck=1 type=rpm-md gpgkey=https://download.ceph.com/keys/release.asc [ceph-source] name=Ceph source package baseurl=http://mirrors.aliyun.com/ceph/rpm-nautilus/el7/SRPMS enabled=1 gpgcheck=1 type=rpm-md gpgkey=https://download.ceph.com/keys/release.asc ``` for i in deploy node1 node2 node3;do scp /root/ceph.repo root@$i:/etc/yum.repos.d/";done for i in deploy node1 node2 node3;do scp /root/centos.repo root@$i:/etc/yum.repos.d/;done for i in deploy node1 node2 node3;do ssh $i "yum clean all";done
yum install -y epel-release
yum install -y ceph-deploy
1、所有节点安装ceph包（node1 node2 node3）
for i in node1 node2 node3;do ssh $i "yum install -y epel-release";done for i in node1 node2 node3;do ssh $i "yum install -y ceph ceph-radosgw";done
2、Create the cluster
[root@deploy ~]# mkdir /etc/ceph [root@deploy ~]# ls [root@deploy ~]# cd /etc/ceph [root@deploy ceph]# ls [root@deploy ceph]# ceph-deploy new node1 node2 node3 [root@deploy ceph]# ceph-deploy mon create-initial [root@deploy ceph]# ceph-deploy osd create node1 --data /dev/sdb [root@deploy ceph]# ceph-deploy osd create node2 --data /dev/sdb [root@deploy ceph]# ceph-deploy osd create node3--data /dev/sdb root@deploy ceph]# ceph-deploy mgr create node1 node2 node3
[root@deploy ceph]# ceph-deploy admin node1 node2 node3 #推送key
[root@deploy ceph]# ceph-deploy osd prepare node1:/dev/sdb
于是，网上又找方案，应该是conf不同步所致，几圈下来，--overwrite的使用各不相同
root@deploy ceph]# ceph-deploy --overwrite-conf config push deploy node1 node2 node3
原文：https://www.cnblogs.com/deny/p/14774419.html