手动部署CEPH rbd

手动部署 CEPH rbd

在一台机器上进行验证使用：同一台机器上部署一个ceph-mon和一个ceph-osd。

首先查看下系统内核是否支持rbd，如果有错误提示，说明内核不支持，需要升级内核
# modprobe rbd
FATAL: Module rbd not found.

尴尬不，没有rbd内核，其实我是部署完之后才发现的。升级下内核。

#modprobe rbd

--------------------
All-nodes: 如果是集群就在所有机器上进行设置
--------------------

1- 各节点设置 hosts
# vim /etc/hosts

10.200.80.18 node18

...

--------------------
Monitor:在监视器节点上进行，集群的话需要修改配置文件，将所有monitor写入配置文件即可
--------------------

1- 确定 fsid
# uuidgen
1c2f9d3b-fc18-4753-8651-39ca6223fe64

2- 配置文件
# vim /etc/ceph/ceph.conf
[global]
fsid=1c2f9d3b-fc18-4753-8651-39ca6223fe64
mon initial members = node18
mon host = 10.200.80.18
public network = 10.200.80.0/24
auth cluster required = cephx
auth service required = cephx
auth client required = cephx
osd journal size = 1024
filestore xattr use omap = true
osd pool default size = 1
osd pool default min size = 1
osd pool default pg num = 333
osd pool default pgp num = 333
osd crush chooseleaf type = 1
rbd_default_features = 1
[mon]
        mon host = node18
        mon addr = 10.200.80.186789
[mon.a]
        host = node18
        mon addr = 10.200.80.18:6789
[osd.0]
        host = node18

3- 创建 keyring

# ceph-authtool --create-keyring /tmp/ceph.mon.keyring --gen-key -n mon. --cap mon 'allow *'

4- 生成管理员 keyring

# ceph-authtool --create-keyring /etc/ceph/ceph.client.admin.keyring --gen-key -n client.admin --set-uid=0 --cap mon 'allow *' --cap osd 'allow *' --cap mds 'allow'

5- 将 client.admin key 导入 ceph.mon.keyring

# ceph-authtool /tmp/ceph.mon.keyring --import-keyring /etc/ceph/ceph.client.admin.keyring

6- 生成 monitor map

# monmaptool --create --add node18 10.200.80.18 --fsid 0a933b1b-6a20-4627-91a1-36adb6a0d709 /tmp/monmap

7- 创建 monitor 数据目录 ceph-{hostname}

# mkdir /var/lib/ceph/mon/ceph-node18

8- 导入 monitor map 和 keyring 信息

# ceph-mon --mkfs -i node18 --monmap /tmp/monmap --keyring /tmp/ceph.mon.keyring

9- 创建两个空文件

# touch /var/lib/ceph/mon/ceph-node18/done

# touch /var/lib/ceph/mon/ceph-node18/sysvinit

10- 启动 monitor

# service ceph start mon.node18

=== mon.node18 ===
Starting Ceph mon.node18 on node18...
Running as unit run-5055.service.
Starting ceph-create-keys on node18...

上面方法不到能启动就使用下面命令
#ceph-mon -i node18

11- 检查一下
# ceph -s

cluster 0a933b1b-6a20-4627-91a1-36adb6a0d709
health HEALTH_ERR
64 pgs stuck inactive
64 pgs stuck unclean
no osds
monmap e1: 1 mons at {node18=10.200.80.18:6789/0}
election epoch 2, quorum 0 node18
osdmap e1: 0 osds: 0 up, 0 in
pgmap v2: 64 pgs, 1 pools, 0 bytes data, 0 objects
0 kB used, 0 kB / 0 kB avail
64 creating

还没有osd，health为ERR。

#ceph --cluster=ceph --admin-daemon /var/run/ceph/ceph-mon.node18.asok mon_status
{ "name": "node18",
  "rank": 0,
  "state": "leader",
  "election_epoch": 1,
  "quorum": [
        0],
  "outside_quorum": [],
  "extra_probe_peers": [],
  "sync_provider": [],
  "monmap": { "epoch": 1,
      "fsid": "1c2f9d3b-fc18-4753-8651-39ca6223fe64",
      "modified": "2018-02-06 10:49:13.819026",
      "created": "2018-02-06 10:49:13.819026",
      "mons": [
            { "rank": 0,
              "name": "node18",
              "addr": "10.200.80.18:6789\/0"}]}}


12- 将 keyring 和 ceph.conf 拷贝到各个 osd 节点 比如10.200.80.19

# scp /etc/ceph/ceph.conf 10.200.80.19:/etc/ceph/
# scp /etc/ceph/ceph.client.admin.keyring 10.200.80.19:/etc/ceph/


--------------------
Osds：每个机器可以多个osd，每块硬盘或分区可以单独做一个osd。osd个数按实际启动的osd进程计算。
--------------------

1- Create OSD

# uuidgen

d6354bb5-b390-44b5-948f-fbb7f939e53a

# ceph osd create d6354bb5-b390-44b5-948f-fbb7f939e53a
0

2- 创建数据存储目录

--目录方式
# mkdir -p /data/ceph/osd/ceph-0
# ln -s /data/ceph/osd/ceph-0 /var/lib/ceph/osd/ceph-0


--磁盘方式
  存储挂的磁阵，这里要把挂的硬盘分区、格式化
# fdisk /dev/vdb
# mkfs.ext4 /dev/vdb1

  挂载
# mount -o defaults,_netdev /dev/vdb1 /var/lib/ceph/osd/ceph-0

  写入分区表
# vim /etc/fstab
/dev/vdb1 /var/lib/ceph/osd/ceph-0 ext4 defaults,_netdev 0 0


3- 初始化 OSD 数据目录


# ceph-osd -i 0 --mkfs --mkjournal --mkkey \
> --osd-uuid d6354bb5-b390-44b5-948f-fbb7f939e53a \
> --cluster ceph \
> --osd-data=/data/ceph/osd/ceph-0 \
> --osd-journal=/data/ceph/osd/ceph-0/journal

4- 注册 OSD authentication key

# ceph auth add osd.0 osd 'allow *' mon 'allow profile osd' -i /data/ceph/osd/ceph-0/keyring
added key for osd.0

5- 将 ceph 节点加入 CRUSH map

# ceph osd crush add-bucket node6 host
added bucket node6 type host to crush map

# ceph osd crush move node6 root=default
moved item id -2 name 'node6' to location {root=default} in crush map


6- 将 OSD 加入 CRUSH map 此处设置权重为1.0

# ceph osd crush add osd.0 1.0 host=node6
add item id 0 name 'osd.0' weight 1 at location {host=node6} to crush map


7- 创建初始化文件


# touch /var/lib/ceph/osd/ceph-0/sysvinit

8- 启动服务

# service ceph start osd.0
=== osd.0 ===
create-or-move updated item name 'osd.0' weight 0.96 at location {host=node6,root=default} to crush map
Starting Ceph osd.0 on node6...
Running as unit run-27621.service.


如果上述命令不成功,使用下面命令
#ceph-osd -i 0 
starting osd.0 at :/0 osd_data /var/lib/ceph/osd/ceph-0 /var/lib/ceph/osd/ceph-0/journal

9- 检查

# ceph -s
    cluster 1c2f9d3b-fc18-4753-8651-39ca6223fe64
     health HEALTH_WARN mon.node18 low disk space
     monmap e1: 1 mons at {node18=10.200.80.18:6789/0}, election epoch 1, quorum 0 node18
     osdmap e23: 1 osds: 1 up, 1 in
      pgmap v790: 525 pgs, 4 pools, 12052 kB data, 15 objects
            63788 MB used, 100 GB / 171 GB avail
                 525 active+clean


# ceph osd tree
# id weight type name up/down reweight
-1 1 root default
-2 1 host node18
0 1 osd.0 up 1