环境描述：ceph1和ceph2两个服务器（其中ceph1已经安装好了ceph-deploy）

                 每个服务器上各一块S3500 SSD、两块日立1T的HDD盘

                 前后端和管理网络合一：分别为128.128.128.9、128.128.128.10

1、在/etc/hosts中增加：

128.128.128.9  ceph1128.128.128.10 ceph2

2、通过

ssh-keygen和ssh-copy-id

使节点之间免密

3、在ceph1中新建一个目录，保留配置文件

mkdir /root/myclustercd /root/mycluster

4、安装ceph

ceph-deploy install ceph1 ceph2

5、安装mon

ceph-deploy new ceph1

6、修改ceph.conf配置文件

public_network = 128.128.128.0/24cluster_network = 128.128.128.0/24enable experimental unrecoverable data corrupting features = bluestore rocksdb debug_white_box_testing_ec_overwritesbluestore block db size = 10737418240  #10Gbluestore block wal size = 10737418240  #10Gosd objectstore = bluestoremon_allow_pool_delete = truerbd_cache = false[osd]bluestore = true

7、初始化mon

ceph-deploy mon create-initial

若需要增加mon

ceph-deploy add mon ceph2

8、格式化分区

ceph-deploy zap {ceph-node}:{dest-disk}例如：ceph-deploy zap ceph1:/dev/sdb

9、将ceph.bootstrap-osd.keyring拷贝到/var/lib/ceph/bootstrap-osd/下，并命令为ceph.keyring

cp /home/mycluster/ceph.bootstrap-osd.keyring /var/lib/ceph/bootstrap-osd/ceph.keyring或scp /home/mycluster/ceph.bootstrap-osd.keyring ceph2：/var/lib/ceph/bootstrap-osd/ceph.keyring

10、增加osd

ceph-disk prepare  --bluestore --block.db /dev/sdb  --block.wal /dev/sdb /dev/sdc其中：/dev/sdb为SSD盘，/dev/sdc为HDD盘