温馨提示×

温馨提示×

您好,登录后才能下订单哦!

密码登录×
登录注册×
其他方式登录
点击 登录注册 即表示同意《亿速云用户服务条款》

Ceph手动添加osd的步骤

发布时间:2021-07-30 17:20:18 来源:亿速云 阅读:446 作者:chen 栏目:云计算

这篇文章主要介绍“Ceph手动添加osd的步骤”,在日常操作中,相信很多人在Ceph手动添加osd的步骤问题上存在疑惑,小编查阅了各式资料,整理出简单好用的操作方法,希望对大家解答”Ceph手动添加osd的步骤”的疑惑有所帮助!接下来,请跟着小编一起来学习吧!

1、Ceph版本

ceph手动添加osd的过程,其实就是ceph-deploy的过程自己手动执行了一遍

# ceph version
ceph version 12.2.7 (3ec878d1e53e1aeb47a9f619c49d9e7c0aa384d5) luminous (stable)

2、磁盘分区

bluestore存储方式,data是以前的journal(vdc1),block(vdc2)才是实际数据存储

2.1、划分ceph data

# uuidgen 
ff3db0d3-fd32-4b2d-8c35-1fb074e00cea
# sgdisk --new=1:0:+100M --change-name=1:"ceph data" --partition-guid=1:ff3db0d3-fd32-4b2d-8c35-1fb074e00cea --typecode=1:89c57f98-2fe5-4dc0-89c1-f3ad0ceff2be --mbrtogpt -- /dev/vdc
The operation has completed successfully.
# /usr/bin/udevadm settle --timeout=600
# /usr/bin/flock -s /dev/vdc /usr/sbin/partprobe /dev/vdc
# /usr/bin/udevadm settle --timeout=600

2.2、划分ceph block

# uuidgen 
a44651fb-8904-4a86-adf6-541fefdf229e
# sgdisk --largest-new=2 --change-name=2:"ceph block" --partition-guid=2:a44651fb-8904-4a86-adf6-541fefdf229e --typecode=2:cafecafe-9b03-4f30-b4c6-b4b80ceff106 --mbrtogpt -- /dev/vdc
The operation has completed successfully.
# /usr/bin/udevadm settle --timeout=600
# /usr/bin/flock -s /dev/vdc /usr/sbin/partprobe /dev/vdc
# /usr/bin/udevadm settle --timeout=600

2.3、查看分区情况

# ls -lrt /dev/disk/by-partuuid/|grep vdc
a44651fb-8904-4a86-adf6-541fefdf229e -> ../../vdc2
lrwxrwxrwx 1 root root 10 Jul 12 16:37
ff3db0d3-fd32-4b2d-8c35-1fb074e00cea -> ../../vdc1

3、格式化/dev/vdc1为xfs

# sgdisk --typecode=2:cafecafe-9b03-4f30-b4c6-b4b80ceff106 -- /dev/vdc
The operation has completed successfully.
# udevadm settle --timeout=600
# flock -s /dev/vdc /usr/sbin/partprobe /dev/vdc
# udevadm settle --timeout=600
# mkfs -t xfs -f -i size=2048 -- /dev/vdc1
meta-data=/dev/vdc1              isize=2048   agcount=4, agsize=6400 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=1        finobt=0, sparse=0
data     =                       bsize=4096   blocks=25600, imaxpct=25
         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
log      =internal log           bsize=4096   blocks=864, version=2
         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0

4、挂载临时目录(vdc1)

# mkdir /var/lib/ceph/tmp/mnt.DlWdC2
# mount -t xfs -o noatime,inode64 -- /dev/vdc1 /var/lib/ceph/tmp/mnt.DlWdC2

# fsid=$(ceph-osd --cluster=ceph --show-config-value=fsid)
# cat << EOF > /var/lib/ceph/tmp/mnt.DlWdC2/ceph_fsid
> $fsid
> EOF
# echo "ff3db0d3-fd32-4b2d-8c35-1fb074e00cea" >> /var/lib/ceph/tmp/mnt.DlWdC2/fsid
# restorecon -R /var/lib/ceph/tmp/mnt.DlWdC2/magic 
# cat << EOF > /var/lib/ceph/tmp/mnt.DlWdC2/block_uuid
> a44651fb-8904-4a86-adf6-541fefdf229e
> EOF

5、建立block软链接(vdc2)

# ln -s /dev/disk/by-partuuid/a44651fb-8904-4a86-adf6-541fefdf229e /var/lib/ceph/tmp/mnt.DlWdC2/block
# echo bluestore >> /var/lib/ceph/tmp/mnt.DlWdC2/type
# /usr/sbin/restorecon -R /var/lib/ceph/tmp/mnt.DlWdC2/
# /usr/bin/chown -R ceph:ceph /var/lib/ceph/tmp/mnt.DlWdC2/
# umount /var/lib/ceph/tmp/mnt.DlWdC2/
# sgdisk --typecode=1:4fbd7e29-9d25-41b8-afd0-062c0ceff05d -- /dev/vdc
Warning: The kernel is still using the old partition table.
The new table will be used at the next reboot.
The operation has completed successfully.

6、启动osd

udev来实现了osd的自动挂载

# /usr/bin/udevadm settle --timeout=600
# /usr/bin/flock -s /dev/vdc /usr/sbin/partprobe /dev/vdc
# /usr/bin/udevadm settle --timeout=600
# /usr/bin/udevadm trigger --action=add --sysname-match vdc1

7、osd目录文件说明

# cd /var/lib/ceph/osd/ceph-4/
# ls -lrt
total 52
-rw-r--r-- 1 ceph ceph  37 Jul 12 16:43 ceph_fsid
-rw-r--r-- 1 ceph ceph  37 Jul 12 16:47 fsid
-rw-r--r-- 1 ceph ceph  21 Jul 12 16:47 magic
-rw-r--r-- 1 ceph ceph  37 Jul 12 16:49 block_uuid
lrwxrwxrwx 1 ceph ceph  58 Jul 12 16:50 block -> /dev/disk/by-partuuid/a44651fb-8904-4a86-adf6-541fefdf229e
-rw-r--r-- 1 ceph ceph  10 Jul 12 16:51 type
-rw------- 1 ceph ceph  56 Jul 12 16:57 keyring
-rw-r--r-- 1 ceph ceph   2 Jul 12 16:57 whoami
-rw-r--r-- 1 root root 384 Jul 12 16:57 activate.monmap
-rw-r--r-- 1 ceph ceph   8 Jul 12 16:57 kv_backend
-rw-r--r-- 1 ceph ceph   2 Jul 12 16:57 bluefs
-rw-r--r-- 1 ceph ceph   4 Jul 12 16:57 mkfs_done
-rw-r--r-- 1 ceph ceph   6 Jul 12 16:57 ready
-rw-r--r-- 1 ceph ceph   3 Jul 12 16:57 active
-rw-r--r-- 1 ceph ceph   0 Jul 12 16:57 systemd
ceph_fsid       -- ceph集群的fsid
fsid            -- osd的fsid,也是/dev/vdc1的盘符id
magic           -- ceph osd volume v026
block_uuid      -- /dev/vdc2的盘符id
block           -- 软链接,指向/dev/vdc2的盘符
type            -- ceph存储类型,这里为bluestore
keyring         -- osd的秘钥
whoami          -- osd的序号
activate.monmap -- 活跃的monmap
kv_backend      -- 值为rocksdb
bluefs          -- 值为1
mkfs_done       -- 值为yes
ready           -- 值为ready
active          -- 值为ok
systemd         -- 空文件

到此,关于“Ceph手动添加osd的步骤”的学习就结束了,希望能够解决大家的疑惑。理论与实践的搭配能更好的帮助大家学习,快去试试吧!若想继续学习更多相关知识,请继续关注亿速云网站,小编会继续努力为大家带来更多实用的文章!

向AI问一下细节

免责声明:本站发布的内容(图片、视频和文字)以原创、转载和分享为主,文章观点不代表本网站立场,如果涉及侵权请联系站长邮箱:is@yisu.com进行举报,并提供相关证据,一经查实,将立刻删除涉嫌侵权内容。

AI