本手册详解讲解部署、运维和使用 Ceph 的过程。
部署：涉及 Ceph 资源规划、组件安装&配置、状态检查等，提供一个高性能、高可靠性、多功能的存储集群；
运维：扩容、下线节点、常见问题和故障、Troubleshooting 等；
应用：详细演示 磁盘快、对象、文件系统 的使用方式，以及作为 K8S 持久化存储的使用方式(PV PVC StorageClass) 等；
一、部署
本手册讲解使用 ceph-deploy 工具部署 luminous 版本 Ceph 集群的步骤。
主机规划如下：
| IP | 主机名 | 内容 |
|---|---|---|
| 172.27.132.65 | kube-node1 | mgr、mon、osd |
| 172.27.132.66 | kube-node2 | osd |
| 172.27.132.67 | kube-node3 | mds、osd |
1. 节点初始化
节点初始化
配置软件源
sudo yum install -y epel-release
cat << EOM > /etc/yum.repos.d/ceph.repo
[ceph-noarch]
name=Ceph noarch packages
baseurl=https://download.ceph.com/rpm-luminous/el7/noarch
enabled=1
gpgcheck=1
type=rpm-md
gpgkey=https://download.ceph.com/keys/release.asc
EOM
安装依赖的软件包
sudo yum install -y ntp ntpdate ntp-doc openssh-server
创建和配置 ceph 账户
在所有的 Ceph Node 创建 ceph 运行的专有账户：
sudo useradd -d /home/ceph -m ceph
sudo passwd ceph # 这里设置密码为 ceph
为 Ceph 用户添加 sudo 权限：
echo “ceph ALL = (root) NOPASSWD:ALL” | sudo tee /etc/sudoers.d/ceph
sudo chmod 0440 /etc/sudoers.d/ceph
配置主机别名
在所有节点上设置 hosts，使得各个 ceph node 可以通过 hostname 访问，注意访问自己的 hostname 的时候，不能解析到 127.0.0.1:
$ grep node /etc/hosts
172.27.132.65 kube-node1 kube-node1
172.27.132.66 kube-node2 kube-node2
172.27.132.67 kube-node3 kube-node3
关闭 SELinux
关闭 SELinux，否则后续 K8S 挂载目录时可能报错 Permission denied：
$ sudo setenforce 0
$ grep SELINUX /etc/selinux/config
SELINUX=disabled
修改配置文件，永久生效；
其它
关闭 requiretty：修改 /etc/sudoers 文件，注释 Defaults requiretty，或设置为：Defaults:ceph !requiretty
初始化 Ceph deploy 节点
按照规则，172.27.132.65 kube-node1 节点将作为 deploy 节点。
配置 kube-node1 节点的 ceph 账户可以免密码登陆到所有节点(包括自身)：
su -l ceph
ssh-keygen -t rsa
ssh-copy-id [email protected]
ssh-copy-id [email protected]
ssh-copy-id [email protected]
设置 kube-node1 的 ceph 账户默认登录其它节点的用户名为 ceph：
cat >>/home/ceph/.ssh/config <<EOF
Host kube-node1
Hostname kube-node1
User ceph
Host kube-node2
Hostname kube-node2
User ceph
Host kube-node3
Hostname kube-node3
User ceph
EOF
chmod 600 ~/.ssh/config
安装 ceph-deploy 工具：
sudo yum update
sudo yum install ceph-deploy
2. 部署 monitor 节点
创建 Ceph 集群和部署 monitor 节点
如果未指明，本文档中的所有操作均在 deploy 节点 ceph 用户家目录 (/home/ceph) 下进行。
创建 deploy 工作目录，保存安装过程中生成的文件:
su -l ceph
mkdir my-cluster
cd my-cluster
创建 ceph 集群
创建名为 ceph 的集群：
[[email protected] my-cluster]$ ceph-deploy new kube-node1 # 参数为初始的 monitor 节点（实际上只是在当前目录下生成 ceph.conf 和 ceph.mon.keyring 文件）
输出：
[ceph_deploy.conf][DEBUG ] found configuration file at: /home/ceph/.cephdeploy.conf
[ceph_deploy.cli][INFO ] Invoked (2.0.0): /bin/ceph-deploy new kube-node1
…
[ceph_deploy.new][DEBUG ] Creating new cluster named ceph
[ceph_deploy.new][INFO ] making sure passwordless SSH succeeds
[kube-node1][DEBUG ] connection detected need for sudo
[kube-node1][DEBUG ] connected to host: kube-node1
[kube-node1][DEBUG ] detect platform information from remote host
[kube-node1][DEBUG ] detect machine type
[kube-node1][DEBUG ] find the location of an executable
[kube-node1][INFO ] Running command: sudo /usr/sbin/ip link show
[kube-node1][INFO ] Running command: sudo /usr/sbin/ip addr show
[kube-node1][DEBUG ] IP addresses found: [u’172.30.53.0’, u’172.30.53.1’, u’172.27.132.65’]
[ceph_deploy.new][DEBUG ] Resolving host kube-node1
[ceph_deploy.new][DEBUG ] Monitor kube-node1 at 172.27.132.65
[ceph_deploy.new][DEBUG ] Monitor initial members are [‘kube-node1’]
[ceph_deploy.new][DEBUG ] Monitor addrs are [‘172.27.132.65’]
[ceph_deploy.new][DEBUG ] Creating a random mon key…
[ceph_deploy.new][DEBUG ] Writing monitor keyring to ceph.mon.keyring…
[ceph_deploy.new][DEBUG ] Writing initial config to ceph.conf…
命令结束后，在当前工作目录自动生成了集群配置文件 ceph.conf、日志文件和用于 bootstrap monitor 节点的 ceph.mon.keyring 文件：
[[email protected] my-cluster]$ ls .
ceph.conf ceph-deploy-ceph.log ceph.mon.keyring
修改 ceph.conf 文件中的默认配置，最终结果如下：
[[email protected] my-cluster]$ cat ceph.conf
[global]
fsid = 0dca8efc-5444-4fa0-88a8-2c0751b47d28
初始 monitor 节点
mon_initial_members = kube-node1
mon_host = 10.64.3.9
cephx 认证授权
auth_cluster_required = cephx
auth_service_required = cephx
auth_client_required = cephx
副本数，应该 <= OSD 数目
osd pool default size = 3
最小副本数
osd pool default min size = 1
PG 和 PGP 默认值
osd pool default pg num = 128
osd pool default pgp num = 128
只启用 Centos Kernel 支持的 layering 特性
rbd_default_features = 1
osd crush chooseleaf type = 1
max mds = 5
mds max file size = 100000000000000
mds cache size = 1000000
文件系统调优
osd_mkfs_type = xfs
osd_mount_options_xfs = rw,noatime,inode64,logbsize=256k,delaylog
osd_mkfs_options_xfs = -f -i size=2048
Journal 调优
journal_max_write_entries = 1000
journal_queue_max_ops = 3000
journal_max_write_bytes = 1048576000
journal_queue_max_bytes = 1048576000
Op tracker
osd_enable_op_tracker = false
OSD Client
osd_client_message_size_cap = 0
osd_client_message_cap = 0
Objector
objecter_inflight_ops = 102400
objector_inflight_op_bytes = 1048576000
Throttles
ms_dispatch_throttle_bytes = 1048576000
OSD Threads
osd_op_threads = 32
osd_op_num_shards = 5
osd_op_num_threads_per_shard = 2
Network，适用于多个网卡的情况
public network = 10.0.0.0/8 # 适用于 ceph client 与集群间的通信
cluster network = 10.0.0.0/8 # 适用于 ceph OSD 之间的数据传输和通信
在所有节点上安装 ceph 软件包（ceph 和 ceph-radosgw）：
–release 参数指定安装的版本为 luminous（不指定时默认为 jewel）：
[[email protected] my-cluster]$ ceph-deploy install --release luminous kube-node1 kube-node2 kube-node3
部署 monitor 节点
初始化 ceph-deploy new kube-node1 命令指定的初始 monitor 节点：
[[email protected] my-cluster]$ ceph-deploy mon create-initial # create-initial/stat/remove
输出：
kube-node1][INFO ] monitor: mon.kube-node1 is running
[kube-node1][INFO ] Running command: sudo ceph --cluster=ceph --admin-daemon /var/run/ceph/ceph-mon.kube-node1.asok mon_status
[ceph_deploy.mon][INFO ] processing monitor mon.kube-node1
…
[kube-node1][INFO ] Running command: sudo ceph --cluster=ceph --admin-daemon /var/run/ceph/ceph-mon.kube-node1.asok mon_status
[ceph_deploy.mon][INFO ] mon.kube-node1 monitor has reached quorum!
[ceph_deploy.mon][INFO ] all initial monitors are running and have formed quorum
[ceph_deploy.mon][INFO ] Running gatherkeys…
…
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --admin-daemon=/var/run/ceph/ceph-mon.kube-node1.asok mon_status
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-kube-node1/keyring auth get client.admin
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-kube-node1/keyring auth get client.bootstrap-mds
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-kube-node1/keyring auth get client.bootstrap-mgr
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-kube-node1/keyring auth get-or-create client.bootstrap-mgr mon allow profile bootstrap-mgr
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-kube-node1/keyring auth get client.bootstrap-osd
[kube-node1][INFO ] Running command: sudo /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-kube-node1/keyring auth get client.bootstrap-rgw
[ceph_deploy.gatherkeys][INFO ] Storing ceph.client.admin.keyring
[ceph_deploy.gatherkeys][INFO ] Storing ceph.bootstrap-mds.keyring
[ceph_deploy.gatherkeys][INFO ] Storing ceph.bootstrap-mgr.keyring
[ceph_deploy.gatherkeys][INFO ] keyring ‘ceph.mon.keyring’ already exists
[ceph_deploy.gatherkeys][INFO ] Storing ceph.bootstrap-osd.keyring
[ceph_deploy.gatherkeys][INFO ] Storing ceph.bootstrap-rgw.keyring
[ceph_deploy.gatherkeys][INFO ] Destroy temp directory /tmp/tmpw0Skr7
命令执行结束后，会在当前目录生成初始化 mds、osd、rgw 的 keyring 文件 {cluster-name}.bootstrap-{type}.keyring，同时创建了 client.admin 用户和它的 keyring，这些 keyring 都已经保存到集群中，供后续部署相应节点时使用：
[[email protected] my-cluster]$ ls -l
total 476
-rw------- 1 ceph ceph 113 Jul 5 16:08 ceph.bootstrap-mds.keyring
-rw------- 1 ceph ceph 71 Jul 5 16:08 ceph.bootstrap-mgr.keyring
-rw------- 1 ceph ceph 113 Jul 5 16:08 ceph.bootstrap-osd.keyring
-rw------- 1 ceph ceph 113 Jul 5 16:08 ceph.bootstrap-rgw.keyring
-rw------- 1 ceph ceph 129 Jul 5 16:08 ceph.client.admin.keyring
-rw-rw-r-- 1 ceph ceph 201 Jul 5 14:15 ceph.conf
-rw-rw-r-- 1 ceph ceph 456148 Jul 5 16:08 ceph-deploy-ceph.log
-rw------- 1 ceph ceph 73 Jul 5 14:15 ceph.mon.keyring
[[email protected] my-cluster]$ ls -l /var/lib/ceph/
total 0
drwxr-x— 2 ceph ceph 26 Apr 24 00:59 bootstrap-mds
drwxr-x— 2 ceph ceph 26 Apr 24 00:59 bootstrap-mgr
drwxr-x— 2 ceph ceph 26 Apr 24 00:59 bootstrap-osd
drwxr-x— 2 ceph ceph 6 Apr 24 00:59 bootstrap-rbd
drwxr-x— 2 ceph ceph 26 Apr 24 00:59 bootstrap-rgw
drwxr-x— 2 ceph ceph 6 Apr 24 00:59 mds
drwxr-x— 3 ceph ceph 29 Apr 24 00:59 mgr
drwxr-x— 3 ceph ceph 29 Apr 24 00:59 mon
drwxr-x— 2 ceph ceph 6 Apr 24 00:59 osd
drwxr-xr-x 2 root root 6 Apr 24 00:59 radosgw
drwxr-x— 2 ceph ceph 6 Apr 24 00:59 tmp
[[email protected] my-cluster]$ ls /var/lib/ceph//
/var/lib/ceph/bootstrap-mds/ceph.keyring /var/lib/ceph/bootstrap-mgr/ceph.keyring /var/lib/ceph/bootstrap-osd/ceph.keyring /var/lib/ceph/bootstrap-rgw/ceph.keyring
/var/lib/ceph/mon/ceph-kube-node1:
done keyring kv_backend store.db systemd
推送 admin keyring 和 ceph.conf 集群配置文件到所有节点（ /etc/ceph/ 目录下），这样后续执行 ceph 命令时不需要指定 monitor 地址和 ceph.client.admin.keyring 文件路径：
[[email protected] my-cluster]$ ceph-deploy admin kube-node1 kube-node2 kube-node3
部署 manager 节点 (luminous+ 版本才需要部署 manager 节点)
[[email protected] my-cluster] ceph-deploy mgr create kube-node1
输出：
ceph_deploy.conf][DEBUG ] found configuration file at: /home/ceph/.cephdeploy.conf
…
[ceph_deploy.mgr][DEBUG ] Deploying mgr, cluster ceph hosts kube-node1:kube-node1
[kube-node1][DEBUG ] connection detected need for sudo
[kube-node1][DEBUG ] connected to host: kube-node1
[kube-node1][DEBUG ] detect platform information from remote host
[kube-node1][DEBUG ] detect machine type
[ceph_deploy.mgr][INFO ] Distro info: CentOS Linux 7.5.1804 Core
[ceph_deploy.mgr][DEBUG ] remote host will use systemd
[ceph_deploy.mgr][DEBUG ] deploying mgr bootstrap to kube-node1
[kube-node1][DEBUG ] write cluster configuration to /etc/ceph/{cluster}.conf
[kube-node1][INFO ] Running command: sudo ceph --cluster ceph --name client.bootstrap-mgr --keyring /var/lib/ceph/bootstrap-mgr/ceph.keyring auth get-or-create mgr.kube-node1 mon allow profile mgr osd allow * mds allow * -o /var/lib/ceph/mgr/ceph-kube-node1/keyring
[kube-node1][INFO ] Running command: sudo systemctl enable [email protected]
[kube-node1][WARNIN] Created symlink from /etc/systemd/system/ceph-mgr.target.wants/[email protected] to /usr/lib/systemd/system/[email protected]
[kube-node1][INFO ] Running command: sudo systemctl start [email protected]
[kube-node1][INFO ] Running command: sudo systemctl enable ceph.target
查看集群状态
切换到 montor 节点 kube-node1，修改 keyring 文件的权限，使非 root 可读：
$ ssh [email protected]
[[email protected] ~]$ ls /etc/ceph/
ceph.client.admin.keyring ceph.conf rbdmap tmp018nTi
[[email protected] ~]$ ls -l /etc/ceph/ceph.client.admin.keyring
-rw------- 1 root root 129 Mar 11 23:43 /etc/ceph/ceph.client.admin.keyring
[[email protected] ~]$ sudo chmod +r /etc/ceph/ceph.client.admin.keyring
[[email protected] ~]$ ls -l /etc/ceph/ceph.client.admin.keyring
-rw-r–r-- 1 root root 129 Mar 11 23:43 /etc/ceph/ceph.client.admin.keyring
查看当前集群状态：
[[email protected] my-cluster]$ ceph -s
cluster:
id: b7b9e370-ea9b-4cc0-8b09-17167c876c24
health: HEALTH_ERR
64 pgs are stuck inactive for more than 60 seconds
64 pgs stuck inactive
64 pgs stuck unclean
no osds
services:
mon: 1 daemons, quorum kube-node1
mgr: kube-node1(active)
osd: 0 osds: 0 up, 0 in
data:
pools: 1 pools, 64 pgs
objects: 0 objects, 0 bytes
usage: 0 kB used, 0 kB / 0 kB avail
pgs: 100.000% pgs not active
64 creating
查看 monitor 节点信息：
[[email protected] my-cluster]$ ceph mon dump
dumped monmap epoch 2
epoch 2
fsid b7b9e370-ea9b-4cc0-8b09-17167c876c24
last_changed 2018-07-05 16:34:09.194222
created 2018-07-05 16:07:57.975307
0: 172.27.132.65:6789/0 mon.kube-node1
扩容 montor 节点
安装 luminous 版本的 ceph 软件包：
[[email protected] my-cluster]$ ceph-deploy install --release luminous kube-node4
创建新的 mon 节点：
[[email protected] my-cluster]$ ceph-deploy mon create kube-node4
修改 ceph.conf 文件，添加扩容的 kube-node4 节点：
[[email protected] my-cluster]$ cat ceph.conf
…
mon_initial_members = kube-node1,kube-node4
mon_host = 172.27.132.65,172.27.132.68
…
推送新的 ceph.conf 到所有节点：
[[email protected] my-cluster]$ ceph-deploy config push kube-node1 kube-node2 kube-node3 kube-node4
登录所有 monitor 节点，重启 ceph-mon 服务：
[[email protected] my-cluster]$ sudo systemctl restart ‘[email protected]’
删除 montor 节点
停止 montor 服务：
[[email protected] ~]$ sudo systemctl stop ‘[email protected]’
将 monitor kube-node4 从集群删除：
[[email protected] ~]$ ceph mon remove kube-node4
修改 ceph.conf 文件，删除 kube-node3 的 montor 信息：
$ cat ceph.conf
…
mon_initial_members = kube-node1
mon_host = 172.27.132.65
…
推送新的 ceph.conf 到所有节点：
[[email protected] my-cluster]$ ceph-deploy config push kube-node1 kube-node2 kube-node3 kube-node4
登录所有 monitor 节点，重启 ceph-mon 服务：
[[email protected] my-cluster]$ sudo systemctl restart ‘[email protected]’
3. 部署 OSD 节点
部署 OSD 节点
安装 ceph 软件包(ceph 和 ceph-radosgw)，–release 可以指定版本：
[[email protected] my-cluster]$ ceph-deploy install --release luminous kube-node1 kube-node2 kube-node3
[[email protected] my-cluster]$ ceph-deploy config push kube-node1 kube-node2 kube-node3
准备 OSD 数据盘磁盘
OSD 需要使用整块磁盘或分区来保存数据，如果机器已经挂载了数据盘，需要先卸载：
[[email protected] my-cluster]$ df -h |grep /mnt
/dev/vda3 923G 33M 923G 1% /mnt/disk01
[[email protected] my-cluster]$ sudo umount /dev/vda3 # 卸载数据盘分区
检查和卸载所有 OSD 节点的数据盘分区；
部署 OSD 节点
[[email protected] my-cluster]$ ceph-deploy osd create --data /dev/vda3 kube-node1 kube-node2 kube-node3 # --data 指定数据盘分区
输出：
[ceph_deploy.conf][DEBUG ] found configuration file at: /home/ceph/.cephdeploy.conf
…
[ceph_deploy.osd][DEBUG ] Creating OSD on cluster ceph with data device /dev/vda3
[kube-node2][DEBUG ] connection detected need for sudo
[kube-node2][DEBUG ] connected to host: kube-node2
[kube-node2][DEBUG ] detect platform information from remote host
[kube-node2][DEBUG ] detect machine type
[kube-node2][DEBUG ] find the location of an executable
[ceph_deploy.osd][INFO ] Distro info: CentOS Linux 7.3.1611 Core
[ceph_deploy.osd][DEBUG ] Deploying osd to kube-node2
[kube-node2][DEBUG ] write cluster configuration to /etc/ceph/{cluster}.conf
[kube-node2][WARNIN] osd keyring does not exist yet, creating one
[kube-node2][DEBUG ] create a keyring file
[kube-node2][DEBUG ] find the location of an executable
[kube-node2][INFO ] Running command: sudo /usr/sbin/ceph-volume --cluster ceph lvm create --bluestore --data /dev/vda3
[kube-node2][DEBUG ] Running command: /bin/ceph-authtool --gen-print-key
[kube-node2][DEBUG ] Running command: /bin/ceph --cluster ceph --name client.bootstrap-osd --keyring /var/lib/ceph/bootstrap-osd/ceph.keyring -i - osd new 149b2781-077f-4146-82f3-4d8061d24043
[kube-node2][DEBUG ] Running command: vgcreate --force --yes ceph-4f6edf77-637a-4752-888e-df74d102cd4e /dev/vda3
[kube-node2][DEBUG ] stdout: Wiping xfs signature on /dev/vda3.
[kube-node2][DEBUG ] stdout: Physical volume “/dev/vda3” successfully created.
[kube-node2][DEBUG ] stdout: Volume group “ceph-4f6edf77-637a-4752-888e-df74d102cd4e” successfully created
[kube-node2][DEBUG ] Running command: lvcreate --yes -l 100%FREE -n osd-block-149b2781-077f-4146-82f3-4d8061d24043 ceph-4f6edf77-637a-4752-888e-df74d102cd4e
[kube-node2][DEBUG ] stdout: Logical volume “osd-block-149b2781-077f-4146-82f3-4d8061d24043” created.
[kube-node2][DEBUG ] Running command: /bin/ceph-authtool --gen-print-key
[kube-node2][DEBUG ] Running command: mount -t tmpfs tmpfs /var/lib/ceph/osd/ceph-1
[kube-node2][DEBUG ] Running command: chown -R ceph:ceph /dev/dm-0
[kube-node2][DEBUG ] Running command: ln -s /dev/ceph-4f6edf77-637a-4752-888e-df74d102cd4e/osd-block-149b2781-077f-4146-82f3-4d8061d24043 /var/lib/ceph/osd/ceph-1/block
[kube-node2][DEBUG ] Running command: ceph --cluster ceph --name client.bootstrap-osd --keyring /var/lib/ceph/bootstrap-osd/ceph.keyring mon getmap -o /var/lib/ceph/osd/ceph-1/activate.monmap
[kube-node2][DEBUG ] stderr: got monmap epoch 2
[kube-node2][DEBUG ] Running command: ceph-authtool /var/lib/ceph/osd/ceph-1/keyring --create-keyring --name osd.1 --add-key AQD24T1btVJ+MhAAu2BxGrSrv5uhSMvRIzGf3A==
[kube-node2][DEBUG ] stdout: creating /var/lib/ceph/osd/ceph-1/keyring
[kube-node2][DEBUG ] stdout: added entity osd.1 auth auth(auid = 18446744073709551615 key=AQD24T1btVJ+MhAAu2BxGrSrv5uhSMvRIzGf3A== with 0 caps)
[kube-node2][DEBUG ] Running command: chown -R ceph:ceph /var/lib/ceph/osd/ceph-1/keyring
[kube-node2][DEBUG ] Running command: chown -R ceph:ceph /var/lib/ceph/osd/ceph-1/
[kube-node2][DEBUG ] Running command: /bin/ceph-osd --cluster ceph --osd-objectstore bluestore --mkfs -i 1 --monmap /var/lib/ceph/osd/ceph-1/activate.monmap --keyfile - --osd-data /var/lib/ceph/osd/ceph-1/ --osd-uuid 149b2781-077f-4146-82f3-4d8061d24043 --setuser ceph --setgroup ceph
[kube-node2][DEBUG ] --> ceph-volume lvm prepare successful for: /dev/vda3
[kube-node2][DEBUG ] Running command: ceph-bluestore-tool --cluster=ceph prime-osd-dir --dev /dev/ceph-4f6edf77-637a-4752-888e-df74d102cd4e/osd-block-149b2781-077f-4146-82f3-4d8061d24043 --path /var/lib/ceph/osd/ceph-1
[kube-node2][DEBUG ] Running command: ln -snf /dev/ceph-4f6edf77-637a-4752-888e-df74d102cd4e/osd-block-149b2781-077f-4146-82f3-4d8061d24043 /var/lib/ceph/osd/ceph-1/block
[kube-node2][DEBUG ] Running command: chown -R ceph:ceph /dev/dm-0
[kube-node2][DEBUG ] Running command: chown -R ceph:ceph /var/lib/ceph/osd/ceph-1
[kube-node2][DEBUG ] Running command: systemctl enable [email protected]
[kube-node2][DEBUG ] stderr: Created symlink from /etc/systemd/system/multi-user.target.wants/[email protected] to /usr/lib/systemd/system/[email protected]
[kube-node2][DEBUG ] Running command: systemctl start [email protected]
[kube-node2][DEBUG ] --> ceph-volume lvm activate successful for osd ID: 1
[kube-node2][DEBUG ] --> ceph-volume lvm create successful for: /dev/vda3
[kube-node2][INFO ] checking OSD status…
[kube-node2][DEBUG ] find the location of an executable
[kube-node2][INFO ] Running command: sudo /bin/ceph --cluster=ceph osd stat --format=json
[ceph_deploy.osd][DEBUG ] Host kube-node2 is now ready for osd use.
…
调用 ceph-volume --cluster ceph lvm create --bluestore --data /dev/vda3 命令创建 LVM 的 VG 和 LV；
启动 OSD 服务；
查看 OSD 列表
[[email protected] my-cluster]$ ceph-deploy osd list kube-node1
输出：
[ceph_deploy.conf][DEBUG ] found configuration file at: /root/.cephdeploy.conf
[ceph_deploy.cli][INFO ] Invoked (2.0.0): /bin/ceph-deploy osd list kube-node1
[ceph_deploy.cli][INFO ] ceph-deploy options:
[ceph_deploy.cli][INFO ] username : None
[ceph_deploy.cli][INFO ] verbose : False
[ceph_deploy.cli][INFO ] debug : False
[ceph_deploy.cli][INFO ] overwrite_conf : False
[ceph_deploy.cli][INFO ] subcommand : list
[ceph_deploy.cli][INFO ] quiet : False
[ceph_deploy.cli][INFO ] cd_conf : <ceph_deploy.conf.cephdeploy.Conf instance at 0x7f1b37400ab8>
[ceph_deploy.cli][INFO ] cluster : ceph
[ceph_deploy.cli][INFO ] host : [‘kube-node1’]
[ceph_deploy.cli][INFO ] func : <function osd at 0x7f1b37438230>
[ceph_deploy.cli][INFO ] ceph_conf : None
[ceph_deploy.cli][INFO ] default_release : False
[kube-node1][DEBUG ] connected to host: kube-node1
[kube-node1][DEBUG ] detect platform information from remote host
[kube-node1][DEBUG ] detect machine type
[kube-node1][DEBUG ] find the location of an executable
[ceph_deploy.osd][INFO ] Distro info: CentOS Linux 7.5.1804 Core
[ceph_deploy.osd][DEBUG ] Listing disks on kube-node1…
[kube-node1][DEBUG ] find the location of an executable
[kube-node1][INFO ] Running command: /usr/sbin/ceph-volume lvm list
[kube-node1][DEBUG ]
[kube-node1][DEBUG ]
[kube-node1][DEBUG ] ====== osd.0 =======
[kube-node1][DEBUG ]
[kube-node1][DEBUG ] [block] /dev/ceph-7856263d-442b-4e33-9737-b94faa68621b/osd-block-9f987e76-8640-4c3b-a9fd-06f701b63903
[kube-node1][DEBUG ]
[kube-node1][DEBUG ] type block
[kube-node1][DEBUG ] osd id 0
[kube-node1][DEBUG ] cluster fsid b7b9e370-ea9b-4cc0-8b09-17167c876c24
[kube-node1][DEBUG ] cluster name ceph
[kube-node1][DEBUG ] osd fsid 9f987e76-8640-4c3b-a9fd-06f701b63903
[kube-node1][DEBUG ] encrypted 0
[kube-node1][DEBUG ] cephx lockbox secret
[kube-node1][DEBUG ] block uuid bjGqMD-HRZe-vfXP-p2ma-Mofx-j4lK-Gcu0hA
[kube-node1][DEBUG ] block device /dev/ceph-7856263d-442b-4e33-9737-b94faa68621b/osd-block-9f987e76-8640-4c3b-a9fd-06f701b63903
[kube-node1][DEBUG ] vdo 0
[kube-node1][DEBUG ] crush device class None
查看集群 OSD 节点状态
切换到 OSD 节点：
$ ssh kube-node2
查看 OSD 进程和命令行参数：
[[email protected] ~]$ ps -elf|grep ceph-osd|grep -v grep
4 S ceph 23498 1 0 80 0 - 351992 futex_ Jul09 ? 00:08:43 /usr/bin/ceph-osd -f --cluster ceph --id 1 --setuser ceph --setgroup ceph
调整 keryring 文件的读权限，允许非 root 账号读取：
$ ssh [email protected]
[[email protected] ~]$ ls /etc/ceph/
ceph.client.admin.keyring ceph.conf rbdmap tmp018nTi
[[email protected] ~]$ ls -l /etc/ceph/ceph.client.admin.keyring
-rw------- 1 root root 129 Mar 11 23:43 /etc/ceph/ceph.client.admin.keyring
[[email protected] ~]$ sudo chmod +r /etc/ceph/ceph.client.admin.keyring
[[email protected] ~]$ ls -l /etc/ceph/ceph.client.admin.keyring
-rw-r–r-- 1 root root 129 Mar 11 23:43 /etc/ceph/ceph.client.admin.keyring
查看 Ceph 集群状态：
[[email protected] my-cluster]$ ceph -s
cluster:
id: b7b9e370-ea9b-4cc0-8b09-17167c876c24
health: HEALTH_OK
services:
mon: 1 daemons, quorum kube-node1
mgr: kube-node1(active)
osd: 3 osds: 3 up, 3 in
data:
pools: 1 pools, 64 pgs
objects: 0 objects, 0 bytes
usage: 3080 MB used, 2765 GB / 2768 GB avail
pgs: 64 active+clean
查看 OSD 状态：
[[email protected] ~]$ ceph osd stat
3 osds: 3 up, 3 in
查看 OSD 节点树：
[[email protected] ~]$ ceph osd tree
ID CLASS WEIGHT TYPE NAME STATUS REWEIGHT PRI-AFF
-1 2.70419 root default
-9 0 host ambari
-2 0.90140 host kube-node1
0 hdd 0.90140 osd.0 up 1.00000 1.00000
-3 0.90140 host kube-node2
1 hdd 0.90140 osd.1 up 1.00000 1.00000
-4 0.90140 host kube-node3
2 hdd 0.90140 osd.2 up 1.00000 1.00000
Dump OSD 的详细信息：
[[email protected] ~]$ ceph osd dump
epoch 180
fsid b7b9e370-ea9b-4cc0-8b09-17167c876c24
created 2018-07-05 16:07:58.315940
modified 2018-07-10 15:45:17.050188
flags sortbitwise,recovery_deletes,purged_snapdirs
crush_version 10
full_ratio 0.95
backfillfull_ratio 0.9
nearfull_ratio 0.85
require_min_compat_client firefly
min_compat_client firefly
require_osd_release luminous
pool 0 ‘rbd’ replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 64 pgp_num 64 last_change 39 flags hashpspool stripe_width 0 application rbd
removed_snaps [1~3]
max_osd 4
osd.0 up in weight 1 up_from 151 up_thru 177 down_at 148 last_clean_interval [121,147) 172.27.132.65:6800/41937 172.27.132.65:6801/41937 172.27.132.65:6802/41937 172.27.132.65:6803/41937 exists,up 9f987e76-8640-4c3b-a9fd-06f701b63903
osd.1 up in weight 1 up_from 139 up_thru 177 down_at 136 last_clean_interval [129,135) 172.27.132.66:6800/23498 172.27.132.66:6801/23498 172.27.132.66:6802/23498 172.27.132.66:6803/23498 exists,up 149b2781-077f-4146-82f3-4d8061d24043
osd.2 up in weight 1 up_from 175 up_thru 177 down_at 172 last_clean_interval [124,171) 172.27.132.67:6801/28744 172.27.132.67:6802/28744 172.27.132.67:6803/28744 172.27.132.67:6804/28744 exists,up 9aa9379e-a0a9-472f-9c5d-0c36d02c9ebf
测试 OSD 对象存储
[[email protected] ~]$ echo {Test-data} > testfile.txt # 测试文件
[[email protected] ~]$ ceph osd pool create mytest 8 # 创建 pool
[[email protected] ~]$ rbd pool init mytest # 初始化 pool
[[email protected] ~]$ rados put test-object-1 testfile.txt --pool=mytest # 将文件保存到 pool
[[email protected] ~]$ rados -p mytest ls # 查看 pool 中对象列表
test-object-1
[[email protected] ~]$ ceph osd map mytest test-object-1 # 查看对象映射的 PG 和 OSD
osdmap e25 pool ‘mytest’ (1) object ‘test-object-1’ -> pg 1.74dc35e2 (1.2) -> up ([1,2,0], p1) acting ([1,2,0], p1)
[[email protected] ~]$ rados rm test-object-1 --pool=mytest # 删除对象
[[email protected] ~]$ ceph osd pool rm mytest # 删除 pool
扩容 OSD 节点
$ ceph-deploy install kube-node4 # 安装软件包
$ ceph-deploy osd create --data /dev/vda3 kube-node4 # --data 指定数据盘分区，必须先 umount
$ ceph-deploy config push kube-node4 # 将更新的 ceph 配置文件推送到指定的主机列表
删除 OSD 节点
$ sudo systemctl stop ‘[email protected]’ # 为当前 OSD 节点 ID
$ ceph-disk deactivate /dev/vda3
$ ceph-volume lvm zap /dev/vda3
$ ceph-disk destroy /dev/sdb # 从 ceph 中删除 osd
$ ceph osd out # 告诉 mon 这个节点已经不能服务了，需要在其他的osd上进行数据的恢复
$ ceph osd crush remove osd. # 完全从集群的分布当中剔除掉，让集群的crush进行一次重新计算，之前节点还占着这个crush weight，会影响到当前主机的host crush weight
$ ceph auth del osd. # 从认证当中去删除这个节点的信息
$ ceph osd rm # 从集群里面删除这个节点的记录
$ cd my-cluster # 进入 ceph-deploy 的工作目录
$ # 修改 ceph.conf 文件，推送到所有其它节点
$ ceph-deploy --overwrite-conf config push ceph-1 ceph-2 ceph-3
重启 OSD 服务
查看 OSD 进程 ID:
[[email protected] ~]$ ps -elf|grep osd|grep -v grep
4 S ceph 23498 1 0 80 0 - 351992 futex_ Jul09 ? 00:08:47 /usr/bin/ceph-osd -f --cluster ceph --id 1 --setuser ceph --setgroup ceph
根据 ID 重启 OSD 服务：
[[email protected] ~]$ sudo systemctl restart ‘[email protected]’ # @N，N 为 OSD ID；
4. 客户端使用 RBD
客户端使用 RBD
在客户端节点安装 ceph 软件包并拷贝 clien.admin 的 keyring 文件：
[[email protected] my-cluster]$ ceph-deploy install --release luminous kube-node2
[[email protected] my-cluster]$ ceph-deploy config push kube-node2
调整 keryring 文件的读权限，允许非 root 账号读取：
$ ssh [email protected]
[[email protected] ~]$ ls /etc/ceph/
ceph.client.admin.keyring ceph.conf rbdmap tmp018nTi
[[email protected] ~]$ ls -l /etc/ceph/ceph.client.admin.keyring
-rw------- 1 root root 129 Mar 11 23:43 /etc/ceph/ceph.client.admin.keyring
[[email protected] ~]$ sudo chmod +r /etc/ceph/ceph.client.admin.keyring
[[email protected] ~]$ ls -l /etc/ceph/ceph.client.admin.keyring
-rw-r–r-- 1 root root 129 Mar 11 23:43 /etc/ceph/ceph.client.admin.keyring
创建 Pool 和镜像
新建一个 Pool：
[[email protected] ~]$ ceph osd pool create mytest 8
[[email protected] ~]$ rbd pool init mytest
[[email protected] ~]$ ceph osd lspools
0 rbd,1 mytest,
在 Pool 中创建一个 4GB 的镜像：
[[email protected] ~]$ rbd create foo --size 4096 -p mytest
[[email protected] ~]$ rbd list -p mytest
foo
[[email protected] ~]$ rbd info foo -p mytest
rbd image ‘foo’:
size 4096 MB in 1024 objects
order 22 (4096 kB objects)
block_name_prefix: rbd_data.374974b0dc51
format: 2
features: layering
flags:
create_timestamp: Thu Jul 5 17:45:09 2018
使用 RBD 镜像
将 RBD 镜像 foo 映射到本地块设备：
[[email protected] ~]$ sudo rbd map mytest/foo
/dev/rbd0
调整 RBD 磁盘的参数（结果值如下）：
[[email protected] ~]$ cat /sys/block/rbd0/queue/optimal_io_size
4194304
[[email protected] ~]$ cat /sys/block/rbd0/alignment_offset
[[email protected] ~]$ cat /sys/block/rbd0/queue/physical_block_size
512
格式化块设备并挂载：
[[email protected] ~]$ sudo mkfs.ext4 -m0 /dev/rbd0
[[email protected] ~]$ sudo mkdir /mnt/ceph-block-device
[[email protected] ~]$ sudo mount /dev/rbd0 /mnt/ceph-block-device
[[email protected] ~]$ cd /mnt/ceph-block-device
查看本地的映射关系：
[[email protected] ~]$ rbd showmapped
id pool image snap device
0 mytest foo - /dev/rbd0
删除 RBD 镜像
[[email protected] ~]$ umount /dev/rbd0 # umount 磁盘
[[email protected] ~]$ rbd unmap mytest/foo # 删除磁盘映射
[[email protected] ~]$ rbd rm mytest/foo # 删除镜像
Removing image: 100% complete…done.
使用 rbdmap 实现自动 map 和 unmap RBD 磁盘
Client 挂载 RBD Image 后，如果在关机前没有 rbd unmap，则会 hang 在 umount 该 RBD 设备上。
可以通过 rbdmap 解决该问题。rbdmap 是一个 shell 脚本，配置文件：/etc/ceph/rbdmap，格式如下：
[[email protected] ~]$ cat /etc/ceph/rbdmap
# RbdDevice Parameters
#poolname/imagename id=client,keyring=/etc/ceph/ceph.client.keyring
mytest/foo --id admin --keyring /etc/ceph/ceph.client.admin.keyring
5. 部署 RGW 节点
部署 RGW 节点
安装 ceph-radosgw 软件包，–release 可以指定版本：
[[email protected] ~]$ ceph-deploy install --release --rgw kube-node3
输出：
[ceph_deploy.conf][DEBUG ] found configuration file at: /home/ceph/.cephdeploy.conf
[ceph_deploy.cli][INFO ] Invoked (2.0.0): /bin/ceph-deploy install --rgw kube-node2
[kube-node2][WARNIN] altered ceph.repo priorities to contain: priority=1
[kube-node2][INFO ] Running command: sudo yum -y install ceph-radosgw
[kube-node2][DEBUG ] Loaded plugins: fastestmirror, priorities
[kube-node2][WARNIN] Not using downloaded repomd.xml because it is older than what we have:
[kube-node2][WARNIN] Current : Thu Apr 26 21:06:11 2018
[kube-node2][WARNIN] Downloaded: Fri Oct 6 01:41:59 2017
[kube-node2][WARNIN] Not using downloaded repomd.xml because it is older than what we have:
[kube-node2][WARNIN] Current : Thu Apr 26 21:02:10 2018
[kube-node2][WARNIN] Downloaded: Fri Oct 6 01:38:33 2017
[kube-node2][WARNIN] Not using downloaded repomd.xml because it is older than what we have:
[kube-node2][WARNIN] Current : Thu Apr 26 21:02:30 2018
[kube-node2][WARNIN] Downloaded: Fri Oct 6 01:38:39 2017
[kube-node2][DEBUG ] Loading mirror speeds from cached hostfile
[kube-node2][DEBUG ] * epel: mirrors.huaweicloud.com
[kube-node2][DEBUG ] 8 packages excluded due to repository priority protections
[kube-node2][DEBUG ] Package 2:ceph-radosgw-12.2.5-0.el7.x86_64 already installed and latest version
[kube-node2][DEBUG ] Nothing to do
[kube-node2][INFO ] Running command: sudo ceph --version
[kube-node2][DEBUG ] ceph version 12.2.5 (cad919881333ac92274171586c827e01f554a70a) luminous (stable)
配置和启用 RGW 节点
[email protected] my-cluster]$ ceph-deploy rgw create kube-node2 # rgw 默认监听 7480 端口
输出：
[ceph_deploy.conf][DEBUG ] found configuration file at: /home/ceph/.cephdeploy.conf
…
[ceph_deploy.rgw][DEBUG ] Deploying rgw, cluster ceph hosts kube-node2:rgw.kube-node2
[kube-node2][DEBUG ] connection detected need for sudo
[kube-node2][DEBUG ] connected to host: kube-node2
[kube-node2][DEBUG ] detect platform information from remote host
[kube-node2][DEBUG ] detect machine type
[ceph_deploy.rgw][INFO ] Distro info: CentOS Linux 7.3.1611 Core
[ceph_deploy.rgw][DEBUG ] remote host will use systemd
[ceph_deploy.rgw][DEBUG ] deploying rgw bootstrap to kube-node2
[kube-node2][DEBUG ] write cluster configuration to /etc/ceph/{cluster}.conf
[kube-node2][WARNIN] rgw keyring does not exist yet, creating one
[kube-node2][DEBUG ] create a keyring file
[kube-node2][INFO ] Running command: sudo ceph --cluster ceph --name client.bootstrap-rgw --keyring /var/lib/ceph/bootstrap-rgw/ceph.keyring auth get-or-create client.rgw.kube-node2 osd allow rwx mon allow rw -o /var/lib/ceph/radosgw/ceph-rgw.kube-node2/keyring
[kube-node2][INFO ] Running command: sudo systemctl enable [email protected]
[kube-node2][WARNIN] Created symlink from /etc/systemd/system/ceph-radosgw.target.wants/[email protected] to /usr/lib/systemd/system/[email protected]
[kube-node2][INFO ] Running command: sudo systemctl start [email protected]
[kube-node2][INFO ] Running command: sudo systemctl enable ceph.target
[ceph_deploy.rgw][INFO ] The Ceph Object Gateway (RGW) is now running on host kube-node2 and default port 7480
创建和启动 ceph-radosgw 服务；
检查监听的端口
$ ssh [email protected]
[[email protected] ~]# netstat -lnpt|grep radosgw
tcp 0 0 0.0.0.0:7480 0.0.0.0:* LISTEN 33110/radosgw
注意：可以在 ceph.conf 配置文件中修改 RGW 监听的地址，然后重启 rgw 进程生效：
[client]
rgw frontends = civetweb port=80
$ sudo systemctl restart ceph-radosgw.service
测试对象存储
测试端口连通性：
[[email protected] ~]$ curl http://172.27.132.66:7480
<?xml version="1.0" encoding="UTF-8"?>anonymous
创建对象存储账号：
[[email protected] ~]$ radosgw-admin user create --uid=demo --display-name=“ceph sgw demo user”
输出：
{
“user_id”: “demo”,
“display_name”: “ceph sgw demo user”,
“email”: “”,
“suspended”: 0,
“max_buckets”: 1000,
“auid”: 0,
“subusers”: [],
“keys”: [
{
“user”: “demo”,
“access_key”: “BY5C4TRTAH755NH8B8K8”,
“secret_key”: “bdsqOAntwrMJAWTVGngxDPAMXrx7zalSQk8YUwIq”
}
],
“swift_keys”: [],
“caps”: [],
“op_mask”: “read, write, delete”,
“default_placement”: “”,
“placement_tags”: [],
“bucket_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
},
“user_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
},
“temp_url_keys”: [],
“type”: “rgw”
}
s 创建子账号：
[[email protected] ~]$ radosgw-admin subuser create --uid demo --subuser=demo:swift --access=full --secret=secretkey --key-type=swift
输出：
{
“user_id”: “demo”,
“display_name”: “ceph sgw demo user”,
“email”: “”,
“suspended”: 0,
“max_buckets”: 1000,
“auid”: 0,
“subusers”: [
{
“id”: “demo:swift”,
“permissions”: “full-control”
}
],
“keys”: [
{
“user”: “demo”,
“access_key”: “BY5C4TRTAH755NH8B8K8”,
“secret_key”: “bdsqOAntwrMJAWTVGngxDPAMXrx7zalSQk8YUwIq”
}
],
“swift_keys”: [
{
“user”: “demo:swift”,
“secret_key”: “secretkey”
}
],
“caps”: [],
“op_mask”: “read, write, delete”,
“default_placement”: “”,
“placement_tags”: [],
“bucket_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
},
“user_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
},
“temp_url_keys”: [],
“type”: “rgw”
}
生成子账号秘钥：
[[email protected] ~]$ radosgw-admin key create --subuser=demo:swift --key-type=swift --gen-secret
输出：
{
“user_id”: “demo”,
“display_name”: “ceph sgw demo user”,
“email”: “”,
“suspended”: 0,
“max_buckets”: 1000,
“auid”: 0,
“subusers”: [
{
“id”: “demo:swift”,
“permissions”: “full-control”
}
],
“keys”: [
{
“user”: “demo”,
“access_key”: “BY5C4TRTAH755NH8B8K8”,
“secret_key”: “bdsqOAntwrMJAWTVGngxDPAMXrx7zalSQk8YUwIq”
}
],
“swift_keys”: [
{
“user”: “demo:swift”,
“secret_key”: “ttQcU1O17DFQ4I9xzKqwgUe7WIYYX99zhcIfU9vb”
}
],
“caps”: [],
“op_mask”: “read, write, delete”,
“default_placement”: “”,
“placement_tags”: [],
“bucket_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
},
“user_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
},
“temp_url_keys”: [],
“type”: “rgw”
}
测试 S3 bucket
安装库文件
[[email protected] ~]$ sudo yum install python-boto
创建 S3 测试脚本：
import boto.s3.connection
access_key = ‘BY5C4TRTAH755NH8B8K8’
secret_key = ‘bdsqOAntwrMJAWTVGngxDPAMXrx7zalSQk8YUwIq’
conn = boto.connect_s3(
aws_access_key_id=access_key,
aws_secret_access_key=secret_key,
host=‘172.27.132.66’, port=7480,
is_secure=False, calling_format=boto.s3.connection.OrdinaryCallingFormat(),
)
bucket = conn.create_bucket(‘my-new-bucket’)
for bucket in conn.get_all_buckets():
print “{name} {created}”.format(
name=bucket.name,
created=bucket.creation_date,
)
执行 S3 测试脚本：
[[email protected] ~]$ python s3test.py
my-new-bucket 2018-07-11T12:26:05.329Z
测试 swift
安装 swift 客户端：
[[email protected] ~]$ sudo yum install python2-pip
[[email protected] ~]$ sudo pip install python-swiftclient
查看 bucket 列表：
[[email protected] ~]$ swift -V 1.0 -A http://kube-node2:7480/auth -U demo:swift -K ttQcU1O17DFQ4I9xzKqwgUe7WIYYX99zhcIfU9vb list
my-new-bucket
查看 bucket 状态和统计信息：
[[email protected] ~]$ radosgw-admin bucket stats --bucket my-new-bucket
{
“bucket”: “my-new-bucket”,
“zonegroup”: “b71dde3d-8d8d-4240-ad99-1c85182d3e9b”,
“placement_rule”: “default-placement”,
“explicit_placement”: {
“data_pool”: “”,
“data_extra_pool”: “”,
“index_pool”: “”
},
“id”: “1f3f02c4-fe58-4626-992b-c6c0fe4c8acf.64249.1”,
“marker”: “1f3f02c4-fe58-4626-992b-c6c0fe4c8acf.64249.1”,
“index_type”: “Normal”,
“owner”: “demo”,
“ver”: “0#1”,
“master_ver”: “0#0”,
“mtime”: “2018-07-11 20:26:05.332943”,
“max_marker”: “0#”,
“usage”: {},
“bucket_quota”: {
“enabled”: false,
“check_on_raw”: false,
“max_size”: -1,
“max_size_kb”: 0,
“max_objects”: -1
}
}
6. 部署 MDS 节点
部署 metadata 节点
在客户端节点安装 ceph 软件包并拷贝 client.admin 的 keyring 文件：
[[email protected] my-cluster]$ ceph-deploy install --release luminous kube-node3
[[email protected] my-cluster]$ ceph-deploy config push kube-node3
部署 metadata server
[[email protected] my-cluster]$ ceph-deploy mds create kube-node3
检查 meta 服务
登录到部署的 kube-node3 节点，检查服务和端口状态：
$ ssh [email protected]
[[email protected] ~]# systemctl status ‘[email protected]’|grep ‘Active:’
Active: active (running) since 一 2018-07-09 15:45:08 CST; 2 days ago
[[email protected] ~]# netstat -lnpt|grep ceph-mds
tcp 0 0 0.0.0.0:6800 0.0.0.0:* LISTEN 1251/ceph-mds
确保状态为 running，监听端口 6800；
创建 cephfs
创建两个池，最后的数字是 PG 的数量
[[email protected] my-cluster]# ceph osd pool create cephfs_data 100
pool ‘cephfs_data’ created
[[email protected] my-cluster]# ceph osd pool create cephfs_metadata 20
pool ‘cephfs_metadata’ created
创建 cephfs 文件系统，注意一个 ceph 只能创建一个cephfs 文件系统：
[[email protected] my-cluster]# ceph fs new cephfs cephfs_metadata cephfs_data
new fs with metadata pool 3 and data pool 2
创建秘钥文件
查询 client.admin 装好的秘钥：
[[email protected] my-cluster]$ cd ~/my-cluster/
[[email protected] my-cluster]$ grep key ceph.client.admin.keyring
key = AQDe0T1babvPNhAApxQdjXNU20vYqkDG+YWACw==
将上面的秘钥保存到一个文件中，如 admin.secret:
$ cat admin.secret
[client.admin]
key = AQDe0T1babvPNhAApxQdjXNU20vYqkDG+YWACw==
挂载和使用 cephfs
挂载 cephFS 有两种方式：
kernel driver 方式：
[[email protected] ~]$ sudo mkdir /mnt/mycephfs
[[email protected] ~]$ vi admin.secret
[[email protected] ~]$ sudo mount -t ceph kube-node1:6789:/ /mnt/mycephfs -o name=admin,secretfile=admin.secret
1.fuse 方式：
[[email protected] ~]$ sudo yum install ceph-fuse
[[email protected] ~]$ sudo mkdir ~/mycephfs
[[email protected] ~]$ sudo ceph-fuse -k ./ceph.client.admin.keyring -m kube-node1:6789 ~/mycephfs
二、运维
Troubleshooting
-
执行 ceph-deploy new 命令失败，提示 ImportError: No module named pkg_resources
现象：
[[email protected] my-cluster]$ ceph-deploy new kube-node1
Traceback (most recent call last):
File “/bin/ceph-deploy”, line 18, in
from ceph_deploy.cli import main
File “/usr/lib/python2.7/site-packages/ceph_deploy/cli.py”, line 1, in
import pkg_resources
ImportError: No module named pkg_resources
原因:
系统缺少 python2-pip 包。
解决办法：
[[email protected] my-cluster]$ sudo yum install python2-pip
2. 执行 ceph-deploy disk zap 命令失败，提示 AttributeError: ‘Namespace’ object has no attribute ‘debug’
现象：
[[email protected] my-cluster]$ ceph-deploy disk zap kube-node3 /dev/vda3
…
[kube-node3][DEBUG ] find the location of an executable
[ceph_deploy][ERROR ] Traceback (most recent call last):
[ceph_deploy][ERROR ] File “/usr/lib/python2.7/site-packages/ceph_deploy/util/decorators.py”, line 69, in newfunc
[ceph_deploy][ERROR ] return f(*a, **kw)
[ceph_deploy][ERROR ] File “/usr/lib/python2.7/site-packages/ceph_deploy/cli.py”, line 164, in _main
[ceph_deploy][ERROR ] return args.func(args)
[ceph_deploy][ERROR ] File “/usr/lib/python2.7/site-packages/ceph_deploy/osd.py”, line 438, in disk
[ceph_deploy][ERROR ] disk_zap(args)
[ceph_deploy][ERROR ] File “/usr/lib/python2.7/site-packages/ceph_deploy/osd.py”, line 336, in disk_zap
[ceph_deploy][ERROR ] if args.debug:
[ceph_deploy][ERROR ] AttributeError: ‘Namespace’ object has no attribute ‘debug’
原因：
代码 bugs；
解决办法：
sudo vim /usr/lib/python2.7/site-packages/ceph_deploy/osd.py
修改第 336 行
if args.debug:
为
if False:
3. 执行 ceph 命令失败，提示 ERROR: missing keyring, cannot use cephx for authentication
现象：
[[email protected] ~]# sudo ceph-volume lvm create --data /dev/vda3
Running command: /bin/ceph-authtool --gen-print-key
Running command: /bin/ceph --cluster ceph --name client.bootstrap-osd --keyring /var/lib/ceph/bootstrap-osd/ceph.keyring -i - osd new 987317ae-38c1-4236-a621-387d30ff9d36
stderr: 2018-07-05 17:08:49.355424 7fcec57d1700 -1 auth: unable to find a keyring on /var/lib/ceph/bootstrap-osd/ceph.keyring: (2) No such file or directory
stderr: 2018-07-05 17:08:49.355441 7fcec57d1700 -1 monclient: ERROR: missing keyring, cannot use cephx for authentication
stderr: 2018-07-05 17:08:49.355443 7fcec57d1700 0 librados: client.bootstrap-osd initialization error (2) No such file or directory
stderr: [errno 2] error connecting to the cluster
–> RuntimeError: Unable to create a new OSD id
原因：
缺少 /var/lib/ceph/boostrap-XX/*.keyring 文件，这些文件默认位于 deploy 节点上。
解决办法：
在 deploy 节点上执行 ceph-deploy osd create --data /dev/vda3 kube-node2 命令。
-
all OSDs are running luminous or later but require_osd_release < luminous
现象：
[[email protected] my-cluster]$ sudo ceph health
HEALTH_WARN all OSDs are running luminous or later but require_osd_release < luminous
[[email protected] my-cluster]$ ceph -s
cluster:
id: b7b9e370-ea9b-4cc0-8b09-17167c876c24
health: HEALTH_WARN
all OSDs are running luminous or later but require_osd_release < luminous
services:
mon: 1 daemons, quorum kube-node1
mgr: kube-node1(active)
osd: 3 osds: 3 up, 3 in
data:
pools: 1 pools, 64 pgs
objects: 0 objects, 0 bytes
usage: 3079 MB used, 2765 GB / 2768 GB avail
pgs: 64 active+clean
原因：
集群版本信息不一致；
解决办法：
[[email protected] my-cluster]$ ceph osd require-osd-release luminous
recovery_deletes is set
5. create pool 时提示 pg_num 值太大
现象：
[[email protected] ~]$ ceph osd pool create k8s 128 128
Error ERANGE: pg_num 128 size 3 would mean 960 total pgs, which exceeds max 600 (mon_max_pg_per_osd 200 * num_in_osds 3)
原因：
OSD 数目太少，或者 mon_max_pg_per_osd 参数值太小；
解决办法：
扩容 OSD 示例；
调大 mon_max_pg_per_osd 参数值：
[[email protected] my-cluster]$ ceph tell injectargs ‘–mon_max_pg_per_osd=350’
injectargs:mon_max_pg_per_osd = ‘350’ (not observed, change may require restart)
参考：
https://www.wanglf.net/ceph-pg-num-is-too-large.html http://blog.51cto.com/michaelkang/1727667
-
application not enabled on 1 pool(s)
现象：
创建 Pool 后，集群状态异常：
[[email protected] my-cluster]# ceph -s
cluster:
id: b7b9e370-ea9b-4cc0-8b09-17167c876c24
health: HEALTH_WARN
application not enabled on 1 pool(s)
services:
mon: 1 daemons, quorum kube-node1
mgr: kube-node1(active), standbys: kube-node2
osd: 3 osds: 3 up, 3 in
data:
pools: 2 pools, 72 pgs
objects: 1 objects, 12 bytes
usage: 3081 MB used, 2765 GB / 2768 GB avail
pgs: 72 active+clean
原因：
创建的 Pool 没有自动打上 application 的 tag：
或者，Pool 没有初始化；
解决办法：
对创建的 Pool 打 tag：
[[email protected] my-cluster]# ceph osd pool application enable mytest rbd # mytest 为 pool 的名称
enabled application ‘rbd’ on pool ‘mytest’
或者，在创建 Pool 后进行初始化:
ceph osd pool create my-pool 8
ceph pool init my-pool
参考：
https://ceph.com/community/new-luminous-pool-tags/
-
RBD image 映射到本地失败：rbd: sysfs write failed
现象：
映射 RBD 到本地失败：
[[email protected] my-cluster]# sudo rbd map foo -p mytest
rbd: sysfs write failed
RBD image feature set mismatch. Try disabling features unsupported by the kernel with “rbd feature disable”.
In some cases useful info is found in syslog - try “dmesg | tail”.
rbd: map failed: (6) No such device or address
原因：
CentoOS kernel 不支持全部特性：[layering, striping, exclusive-lock, object-map, fast-diff, deep-flatten, journaling, data-pool]
解决方法：
disable RBD 的部分特性：
[[email protected] my-cluster]# rbd feature disable mytest/foo fast-diff,object-map,exclusive-lock,deep-flatten
[[email protected] my-cluster]# rbd info mytest/foo |grep features
features: layering
在 /etc/ceph/ceph.conf 配置文件中添加配置 rbd_default_features = 3，这样创建的 RBD 均只启用 layering 特性：
features 是如下几个值的和：
+1 for layering, +2 for stripingv2, +4 for exclusive lock, +8 for object map +16 for fast-diff, +32 for deep-flatten, +64 for journaling
到 v4.6 linux kernel 只支持 layering 和 stripingv2。
参考：
http://www.zphj1987.com/2016/06/07/rbd%E6%97%A0%E6%B3%95map-rbd-feature-disable/
-
K8S Mount PV 对应的 RBD 镜像失败，提示 rbd: map failed exit status 6 rbd: sysfs write failed
现象：
K8S Mount PV 对应的 RBD 镜像失败，提示 rbd: map failed exit status 6 rbd: sysfs write failed
[[email protected] ~]$ kubectl get pods|grep prometheus-server
wishful-ladybird-prometheus-server-f744b8794-mxr5l 0/2 Init:0/1 0 55s
[[email protected] ~]$ kubectl describe pods wishful-ladybird-prometheus-server-f744b8794-mxr5l|tail -10
Normal Scheduled 2m default-scheduler Successfully assigned wishful-ladybird-prometheus-server-f744b8794-mxr5l to kube-node2
Normal SuccessfulMountVolume 2m kubelet, kube-node2 MountVolume.SetUp succeeded for volume “config-volume”
Normal SuccessfulMountVolume 2m kubelet, kube-node2 MountVolume.SetUp succeeded for volume “wishful-ladybird-prometheus-server-token-8grpr”
Warning FailedMount 59s (x8 over 2m) kubelet, kube-node2 MountVolume.SetUp failed for volume “ceph-pv-8g” : rbd: map failed exit status 6 rbd: sysfs write failed
RBD image feature set mismatch. Try disabling features unsupported by the kernel with “rbd feature disable”.
In some cases useful info is found in syslog - try “dmesg | tail”.
rbd: map failed: (6) No such device or address
Warning FailedMount 2s kubelet, kube-node2 Unable to mount volumes for pod “wishful-ladybird-prometheus-server-f744b8794-mxr5l_default(7cdc39e7-80f0-11e8-9331-525400ce676d)”: timeout expired waiting for volumes to attach/mount for pod “default”/“wishful-ladybird-prometheus-server-f744b8794-mxr5l”. list of unattached/unmounted volumes=[storage-volume]
Warning FailedSync 2s kubelet, kube-node2 Error syncing pod
[[email protected] ~]$ kubectl get pv ceph-pv-8g -o yaml|grep image
image: prometheus-server
直接 disable feature 失败，提示 Read-only file system：
[[email protected] ~]$ rbd feature disable prometheus-server fast-diff,object-map,exclusive-lock,deep-flatten
rbd: failed to update image features: (30) Read-only file system
删除 RBD Image 失败：
[[email protected] ~]$ rbd rm prometheus-server
2018-07-06 18:05:04.968455 7f9ce1ffb700 -1 librbd::image::RemoveRequest: 0x7f9d05abc0f0 handle_exclusive_lock: cannot obtain exclusive lock - not removing
Removing image: 0% complete…failed.
rbd: error: image still has watchers
This means the image is still open or the client using it crashed. Try again after closing/unmapping it or waiting 30s for the crashed client to timeout.
[[email protected] ~]$ rbd status prometheus-server
Watchers: none
原因：
CentoOS kernel 不支持全部特性：[layering, striping, exclusive-lock, object-map, fast-diff, deep-flatten, journaling, data-pool]
解决方法：
删除 RBD Image 的 Lock：
[[email protected] ~]$ rbd lock list prometheus-server
There is 1 exclusive lock on this image.
Locker ID Address
client.44142 kubelet_lock_magic_ambari.hadoop 172.27.132.67:0/710596833
[[email protected] ~]$ rbd lock rm prometheus-server kubelet_lock_magic_ambari.hadoop client.44142
[[email protected] ~]$ rbd rm prometheus-server
Removing image: 100% complete…done.
[[email protected] ~]$
9. unmap rbd 失败
现象：
[[email protected] ~]# sudo rbd unmap foo
rbd: sysfs write failed
rbd: unmap failed: (16) Device or resource busy
原因：
未知；
解决办法：
使用 -o force 强制 unmap:
[[email protected] ~]# sudo rbd unmap -o force foo
参考：
http://lists.ceph.com/pipermail/ceph-users-ceph.com/2017-July/019160.html
-
同时挂载一个 RBD，并发读写后，引起文件系统错误，提示 Input/output error
现象：
[[email protected] ~]# sudo rbd map foo # 第二次挂载同一个 RBD
[[email protected] ~]# sudo mkdir /mnt/ceph-block-device
[[email protected] ~]# sudo mount /dev/rbd0 /mnt/ceph-block-device
[[email protected] ~]# ls /mnt/ceph-block-device/
ls: cannot access /mnt/ceph-block-device/test: Input/output error
lost+found test
[[email protected] ~]# dmesg|tail
[598628.843913] rbd: loaded (major 251)
[598628.850614] libceph: mon0 172.27.132.65:6789 session established
[598628.851312] libceph: client14187 fsid b7b9e370-ea9b-4cc0-8b09-17167c876c24
[598628.860645] rbd: rbd0: capacity 4294967296 features 0x1
[598639.147675] EXT4-fs (rbd0): recovery complete
[598639.147682] EXT4-fs (rbd0): mounted filesystem with ordered data mode. Opts: (null)
[598675.693950] EXT4-fs (rbd0): mounted filesystem with ordered data mode. Opts: (null)
[598678.177275] EXT4-fs error (device rbd0): ext4_lookup:1441: inode #2: comm ls: deleted inode referenced: 12
[598680.603020] EXT4-fs error (device rbd0): ext4_lookup:1441: inode #2: comm ls: deleted inode referenced: 12
[598701.607815] EXT4-fs error (device rbd0): ext4_lookup:1441: inode #2: comm ls: deleted inode referenced: 12
原因：
RBD 不支持多次挂载的并发访问。
解决办法：
使用 fsck 修复。
[[email protected] ~]# umount /dev/rbd1
[[email protected] ~]# fsck -y /dev/rbd1
fsck from util-linux 2.23.2
e2fsck 1.42.9 (28-Dec-2013)
/dev/rbd1 contains a file system with errors, check forced.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Entry ‘test’ in / (2) has deleted/unused inode 12. Clear? yes
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
Inode bitmap differences: -12
Fix? yes
Free inodes count wrong for group #0 (8180, counted=8181).
Fix? yes
Free inodes count wrong (262132, counted=262133).
Fix? yes
/dev/rbd1: ***** FILE SYSTEM WAS MODIFIED *****
/dev/rbd1: 11/262144 files (0.0% non-contiguous), 53326/1048576 blocks
[[email protected] ~]# mount /dev/rbd1 /mnt/ceph-block-device/
[[email protected] ~]# ls -l /mnt/ceph-block-device/
total 16
drwx------ 2 root root 16384 Jul 5 18:07 lost+found
[[email protected] ~]#
11. too few PGs per OSD (16 < min 30)
现象：
$ ceph -s
cluster 85510587-14c6-4526-9636-83179bda2751
health HEALTH_WARN
too few PGs per OSD (16 < min 30)
monmap e3: 3 mons at {controller-01=10.90.3.7:6789/0,controller-02=10.90.3.2:6789/0,controller-03=10.90.3.5:6789/0}
election epoch 8, quorum 0,1,2 controller-02,controller-03,controller-01
osdmap e74: 12 osds: 12 up, 12 in
pgmap v38670: 1408 pgs, 10 pools, 18592 MB data, 4304 objects
56379 MB used, 20760 GB / 20815 GB avail
1408 active+clean
client io 5127 B/s wr, 2 op/s
原因：
Ceph集群目前每个osd的pg数阈值最小是 30，一旦每个osd上的pg数低于30，就会提示“too few PGs per OSD”：
$ ceph --show-config | grep mon_pg_warn_min_per_osd
mon_pg_warn_min_per_osd = 30
解决办法：
查看 Ceph pool
ceph osd lspools 0 rbd,
新建的集群只有一个 rbd 池。
查看rbd 池的PG和PGP值
ceph osd pool get rbd pg_num pg_num: 64
ceph osd pool get rbd pgp_num pgp_num: 64
查看 池的副本数
ceph osd dump | grep size
pool 0 ‘rbd’ replicated size 3 min_size 2 crush_ruleset 0 object_hash rjenkins pg_num 64 pgp_num 64 last_change 1 flags hashpspool stripe_width 0
pgs为64，因为是3副本的配置，所以当有12个osd的时候，每个osd上均分了64/12 *3=16个pgs，低于最低的配置30。
调整 rbd 池的pgs
pg num 的设置不能随便配设置，过大或者过小都不可以。如果过大，那么backfill和recovery的时候负载太大，如果过小，数据就没法很好的均匀分布。 PG总数=（OSD 总数 x100）/ 最大副本数
结果必须舍入到最接近2的N次幂的值。
修改池的PG和PGP：
ceph osd pool set rbd pg_num 512 ceph osd pool set rbd pgp_num 512
不能调整的太大，否则出现 too many PGs per OSD (352 > max 300) 的错误。
-
too many PGs per OSD (352 > max 300)
现象：
$ ceph -s
cluster 85510587-14c6-4526-9636-83179bda2751
health HEALTH_WARN
too many PGs per OSD (352 > max 300)
monmap e3: 3 mons at {controller-01=10.90.3.7:6789/0,controller-02=10.90.3.2:6789/0,controller-03=10.90.3.5:6789/0}
election epoch 8, quorum 0,1,2 controller-02,controller-03,controller-01
osdmap e74: 12 osds: 12 up, 12 in
pgmap v38670: 1408 pgs, 10 pools, 18592 MB data, 4304 objects
56379 MB used, 20760 GB / 20815 GB avail
1408 active+clean
client io 5127 B/s wr, 2 op/s
原因：
Ceph 集群目前每个osd 的 pg 数最大阈值是300，一旦每个osd 的pg数超过300，就会提示“too many PGs per OSD”
$ ceph --show-config | grep mon_pg_warn_max_per_osd
mon_pg_warn_max_per_osd = 300
解决办法：
查看目前的pg分布
pg num和per osd：把所有Pool的PG数量求和，除以OSD的个数，就得到平均每个OSD上的PG数量。
修改所有monitor节点的 ceph.conf
$ vim /etc/ceph/ceph.conf
[global]
mon_pg_warn_max_per_osd = 0 # 在global部分最后添加此行
重启所有的节点
$ restart ceph-mon-all
如果觉得觉得修改配置文件麻烦，也可以直接一条命令搞定：
ceph tell ‘mon.*’ injectargs “–mon_pg_warn_max_per_osd 0”
使用tell命令修改的配置只是临时的，mon 服务重启后会丢失。解决办法是把这个配置加到Ceph mon节点的配置文件里，然后重启mon服务。
-
启动 rgw 失败
现象：
[[email protected] ~]# journalctl -u [email protected]|tail
7月 05 20:04:04 kube-node2 radosgw[29818]: 2018-07-05 20:04:04.799739 7fe6b8908e80 -1 ERROR: failed to initialize watch: (34) Numerical result out of range
7月 05 20:04:04 kube-node2 radosgw[29818]: 2018-07-05 20:04:04.802725 7fe6b8908e80 -1 Couldn’t init storage provider (RADOS)
7月 05 20:04:04 kube-node2 systemd[1]: [email protected]: main process exited, code=exited, status=5/NOTINSTALLED
7月 05 20:04:04 kube-node2 systemd[1]: Unit [email protected] entered failed state.
7月 05 20:04:04 kube-node2 systemd[1]: [email protected] failed.
7月 05 20:04:05 kube-node2 systemd[1]: [email protected] holdoff time over, scheduling restart.
7月 05 20:04:05 kube-node2 systemd[1]: start request repeated too quickly for [email protected]
7月 05 20:04:05 kube-node2 systemd[1]: Failed to start Ceph rados gateway.
7月 05 20:04:05 kube-node2 systemd[1]: Unit [email protected] entered failed state.
7月 05 20:04:05 kube-node2 systemd[1]: [email protected] failed.
原因：
pg_num、pgp_num、mon_max_pg_per_osd 等参数值设置的不正确或太小。
解决办法：
[[email protected] my-cluster]$ ceph tell injectargs ‘–mon_max_pg_per_osd=350’ # 还应该添加到 ceph.conf 文件中；
参考：
http://tracker.ceph.com/issues/22351#note-11
-
s3test.py 返回 416 错误
执行 http://docs.ceph.com/docs/master/install/install-ceph-gateway/ 的 s3 test.py 程序失败：
[[email protected] cert]$ python s3test.py
Traceback (most recent call last):
File “s3test.py”, line 12, in
bucket = conn.create_bucket(‘my-new-bucket’)
File “/usr/lib/python2.7/site-packages/boto/s3/connection.py”, line 625, in create_bucket
response.status, response.reason, body)
boto.exception.S3ResponseError: S3ResponseError: 416 Requested Range Not Satisfiable
原因：
pg_num、pgp_num、mon_max_pg_per_osd 等参数值设置的不正确或太小。
解决办法：
[[email protected] my-cluster]$ ceph tell injectargs ‘–mon_max_pg_per_osd=350’ # 还应该添加到 ceph.conf 文件中；
参考：
https://tracker.ceph.com/issues/21497
-
map rbd 失败，提示 timeout
现象：
$ rbd feature disable foo fast-diff,object-map,exclusive-lock,deep-flatten
$ rbd map foo
rbd: sysfs write failed
In some cases useful info is found in syslog - try “dmesg | tail”.
rbd: map failed: (110) Connection timed out
内核日志：
$ dmesg|tail
[2937489.640621] libceph: mon1 172.27.128.101:6789 feature set mismatch, my 106b84a842a42 < server’s 40106b84a842a42, missing 400000000000000
[2937489.643198] libceph: mon1 172.27.128.101:6789 missing required protocol features
[2937742.427929] libceph: mon2 172.27.128.102:6789 feature set mismatch, my 106b84a842a42 < server’s 40106b84a842a42, missing 400000000000000
[2937742.430234] libceph: mon2 172.27.128.102:6789 missing required protocol features
[2937752.725957] libceph: mon2 172.27.128.102:6789 feature set mismatch, my 106b84a842a42 < server’s 40106b84a842a42, missing 400000000000000
[2937752.728805] libceph: mon2 172.27.128.102:6789 missing required protocol features
[2937762.737960] libceph: mon2 172.27.128.102:6789 feature set mismatch, my 106b84a842a42 < server’s 40106b84a842a42, missing 400000000000000
[2937762.740282] libceph: mon2 172.27.128.102:6789 missing required protocol features
[2937772.722343] libceph: mon2 172.27.128.102:6789 feature set mismatch, my 106b84a842a42 < server’s 40106b84a842a42, missing 400000000000000
[2937772.724659] libceph: mon2 172.27.128.102:6789 missing required protocol features
原因：
Linux kernel version v4.5 开始才支持 400000000000000 对应的 CRUSH_TUNABLES5；
解决办法：
升级内核到 4.5 或更高版本；
或更新 feature flag，不需要 400000000000000；
对于后一种方式，执行命令 ceph osd crush tunables hammer;
参考：
http://cephnotes.ksperis.com/blog/2014/01/21/feature-set-mismatch-error-on-ceph-kernel-client http://lists.ceph.com/pipermail/ceph-users-ceph.com/2017-July/019387.html
-
kuberntes PVC 使用 StorageClass 创建 PV 失败
现象：
$ kubectl get pv # 没有自动创建 PV
No resources found.
$ kubectl describe pvc | tail
Warning ProvisioningFailed 25m persistentvolume-controller Failed to provision volume with StorageClass “ceph”: failed to create rbd image: exit status 1, command output: 2018-07-26 17:35:05.749211 7f4e037687c0 -1 did not load config file, using default settings.
rbd: extraneous parameter --image-feature
$ kubectl get endpoints -n kube-system kube-controller-manager -o yaml|grep leader
control-plane.alpha.kubernetes.io/leader: ‘{“holderIdentity”:“m7-devops-128123”,“leaseDurationSeconds”:15,“acquireTime”:“2018-07-25T14:00:31Z”,“renewTime”:“2018-07-26T10:02:01Z”,“leaderTransitions”:3}’
$ journalctl -u kube-controller-manager |tail -4 # 查看 leader m7-devops-128123 的日志
7月 26 18:01:20 m7-devops-128123 kube-controller-manager[27948]: E0726 18:01:20.798621 27948 rbd.go:367] rbd: create volume failed, err: failed to create rbd image: exit status 1, command output: 2018-07-26 18:01:20.765861 7f912376db00 -1 did not load config file, using default settings.
7月 26 18:01:20 m7-devops-128123 kube-controller-manager[27948]: rbd: extraneous parameter --image-feature
7月 26 18:01:20 m7-devops-128123 kube-controller-manager[27948]: I0726 18:01:20.798671 27948 pv_controller.go:1317] failed to provision volume for claim “default/pvc-test-claim” with StorageClass “ceph”: failed to create rbd image: exit status 1, command output: 2018-07-26 18:01:20.765861 7f912376db00 -1 did not load config file, using default settings.
7月 26 18:01:20 m7-devops-128123 kube-controller-manager[27948]: rbd: extraneous parameter --image-feature
原因：
kuberntes 节点安装的 ceph-common 版本过低，与 ceph 集群不兼容。
解决办法：
配置节点 YUM 源，使用与 ceph 集群版本相匹配的源。然后更新安装 ceph-common。
$ cat /etc/yum.repos.d/ceph.repo
[Ceph]
name=Ceph packages for b a s e a r c h b a s e u r l = h t t p : / / d o w n l o a d . c e p h . c o m / r p m − l u m i n o u s / e l 7 / basearch baseurl=http://download.ceph.com/rpm-luminous/el7/ basearchbaseurl=http://download.ceph.com/rpm−luminous/el7/basearch
enabled=1
gpgcheck=1
type=rpm-md
gpgkey=https://download.ceph.com/keys/release.asc
priority=1
[ceph-noarch]
name=Ceph noarch packages
baseurl=https://download.ceph.com/rpm-luminous/el7/noarch
enabled=1
gpgcheck=1
type=rpm-md
gpgkey=https://download.ceph.com/keys/release.asc
priority=1
[ceph-source]
name=Ceph source packages
baseurl=http://download.ceph.com/rpm-luminous/el7/SRPMS
enabled=1
gpgcheck=1
type=rpm-md
gpgkey=https://download.ceph.com/keys/release.asc
priority=1
$ yum clean all && yum update
17. 创建使用 PVC 的 POD 卡住，一直处于 ContainerCreating 或 Init:0/1 状态（多容器的情况）
现象：
[[email protected] prometheus]# kubectl get pods -n devops-monitoring -o wide|grep -v Running
NAME READY STATUS RESTARTS AGE IP NODE
monitoring-prometheus-alertmanager-67674fb84b-pxfbz 0/2 ContainerCreating 0 16m m7-devops-128107
monitoring-prometheus-server-5777d76b75-l5bdv 0/2 Init:0/1 0 16m m7-devops-128107
prometheus-alertmanager 和 prometheus-server 均使用名为 ceph 的 PVC StorageClass；
POD 一直没有 Running；
查看 StorageClass ceph 详情：
[[email protected] ~]# kubectl get storageclass ceph -o yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
creationTimestamp: 2018-07-26T09:49:30Z
name: ceph
resourceVersion: “462349”
selfLink: /apis/storage.k8s.io/v1/storageclasses/ceph
uid: 3116e9a4-90b9-11e8-b43c-0cc47a2af650
parameters:
adminId: admin
adminSecretName: ceph-secret-admin
adminSecretNamespace: default
imageFeatures: layering
imageFormat: “2”
monitors: 172.27.128.100:6789,172.27.128.101:6789,172.27.128.102:6789
pool: rbd
userId: admin
userSecretName: ceph-secret-admin
provisioner: kubernetes.io/rbd
reclaimPolicy: Delete
该 storageclass 使用的 adminSecret ceph-secret-admin 位于 adminSecretNamespace 指定的 default 命名空间中；
对应节点的 kubelet 日志：
[[email protected] ~]# journalctl -u kubelet -f
– Logs begin at 一 2018-07-30 18:42:12 CST. –
8月 01 17:19:41 m7-devops-128107 kubelet[6120]: E0801 17:19:41.578011 6120 rbd.go:504] failed to get secret from [“devops-monitoring”/“ceph-secret-admin”]
8月 01 17:19:41 m7-devops-128107 kubelet[6120]: E0801 17:19:41.578092 6120 rbd.go:126] Couldn’t get secret from devops-monitoring/&LocalObjectReference{Name:ceph-secret-admin,}
原因：
PVC 指定的 StorageClass 的 userSecretName 必须位于 PVC 所在的命名空间。而 ceph-secret-admin 在 devops-monitoring 命名空间中没有定义，所以出错。
解决办法：
在 PVC 所在的命名空间 devops-monitoring 中创建 ceph-secret-admin secret，注意类型必须是 kubernetes.io/rbd；
[[email protected] k8s]# Secret=$(awk ‘/key = / {print $3}’ /etc/ceph/ceph.client.admin.keyring | base64)
[[email protected] k8s]# cat > ceph-secret-admin.yaml <<EOF
apiVersion: v1
kind: Secret
type: kubernetes.io/rbd
metadata:
name: ceph-secret-admin
namespace: devops-monitoring
data:
key: $Secret
EOF
[[email protected] k8s]# kubectl create -f ceph-secret-admin.yaml
参考： https://kubernetes.io/docs/concepts/storage/storage-classes/#ceph-rbd
![Cinder 配置ceph多后端[图]](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=)