Ubuntu系统用ceph-deploy部署ceph

Posted 2021-08-24 ruyinote

tags:

篇首语：本文由小常识网(cha138.com)小编为大家整理，主要介绍了Ubuntu系统用ceph-deploy部署ceph相关的知识，希望对你有一定的参考价值。

ceph-deploy部署ceph

ceph的组件和功能：

ceph的核心组件包括：OSD、monitor和mds三大组件

OSD：OSD的英文全称是Object Storage Device，它的主要功能是存储数据、复制数据、平衡数据、恢复数据等，与其它OSD间进行心跳检查等，并将一些变化情况上报给Ceph Monitor。一般情况下一块硬盘对应一个OSD，由OSD来对硬盘存储进行管理。

Monitor：由该英文名字我们可以知道它是一个监视器，负责监视Ceph集群，维护Ceph集群的健康状态，同时维护着Ceph集群中的各种Map图，比如OSD Map、Monitor Map、PG Map和CRUSH Map，这些Map统称为Cluster Map，Cluster Map是RADOS的关键数据结构，管理集群中的所有成员、关系、属性等信息以及数据的分发，比如当用户需要存储数据到Ceph集群时，OSD需要先通过Monitor获取最新的Map图，然后根据Map图和object id等计算出数据最终存储的位置。

MDS：全称是Ceph MetaData Server，主要保存的文件系统服务的元数据，但对象存储和块存储设备是不需要使用该服务的。

ceph的数据读写流程

存入的数据先找到要存的pool,把数据分为若干PG，PG默认大小为4M，每个PG生成三个副本（一主两备），放到不同节点的不同osd上。

使用ceph-deploy部署ceph集群

3.1 部署节点信息

节点名称	pubilc network	cluster network	安装系统	部署组件
ceph-deploy	192.168.202.141	192.168.122.141	ubuntu 18.04	ceph-deploy
ubuntu-ceph1	192.168.202.142	192.168.122.142	ubuntu 18.04	monitor,mgr,rgw,mds,osd
ubuntu-ceph2	192.168.202.143	192.168.122.143	ubuntu 18.04	monitor,rgw,mds,osd
ubuntu-ceph3	192.168.202.144	192.168.122.144	ubuntu 18.04	monitor,rgw,mds,osd

ceph版本：ceph version 16.2.5 pacific

deploy版本：ceph-deploy 2.0.1

Ceph网络：Ceph Cluster网络均采用192.168.122.0/24，Ceph Public网络采用192.168.202.0/24

3.2部署ubuntu系统，准备环境（所有节点都执行）

配置IP

root@ubuntu:~# cat /etc/netplan/01-netcfg.yaml

# This file describes the network interfaces available on your system

# For more information, see netplan(5).

network:

version: 2

renderer: networkd

ethernets:

eth0:

dhcp4: no

dhcp6: no

addresses: [192.168.202.141/24]

gateway4: 192.168.202.1

nameservers:

addresses: [114.114.114.114]

eth1:

dhcp4: no

dhcp6: no

addresses: [192.168.122.141/24]

netplan apply

关闭防火墙

ufw status && ufw disable

设置root远程登录

sed -i \'s/#PermitRootLogin prohibit-password/PermitRootLogin yes/g\' /etc/ssh/sshd_config

sed -i \'s/#PasswordAuthentication yes/PasswordAuthentication yes/g\' /etc/ssh/sshd_config

systemctl restart sshd

passwd

配置主机名、hosts文件

hostnamectl set-hostname ceph-deploy

cat << EOF >> /etc/hosts

192.168.202.141 ceph-deploy

192.168.202.142 ubuntu-ceph1

192.168.202.143 ubuntu-ceph2

192.168.202.144 ubuntu-ceph3

EOF

更换源

参考网站 https://www.linuxidc.com/Linux/2018-08/153709.htm

cp /etc/apt/sources.list /etc/apt/sources.list.bak

cat << EOF > /etc/apt/sources.list

deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse

deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse

deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse

deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse

deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse

deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse

deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse

deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse

deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse

deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse

EOF

apt update

安装基础包

apt install iproute2 ntpdate tcpdump telnet traceroute nfs-kernel-server nfs-common lrzsz tree openssl libssl-dev libpcre3 libpcre3-dev zlib1g-dev ntpdate tcpdump telnet traceroute gcc openssh-server lrzsz tree openssl libssl-dev libpcre3 libpcre3-dev zlib1g-dev ntpdate tcpdump telnet traceroute iotop unzip zip

时间同步

apt install chrony -y

systemctl status chronyd

chronyc sources -v

添加ceph源

wget -q -O- \'https://mirrors.tuna.tsinghua.edu.cn/ceph/keys/release.asc\' | sudo apt-key add -

echo "deb https://mirrors.tuna.tsinghua.edu.cn/ceph/debian-pacific bionic main" >> /etc/apt/sources.list

apt-get update

创建部署用户及配置sudo权限

a.考虑到使用root用户的安全性问题，所以这里创建一个 ceph-admin 普通用户做为部署及运维使用

b.再加上ceph-deploy会在节点安装软件包，所以创建的用户需要无密码 sudo 权限

groupadd -r -g 2022 ceph-admin && useradd -r -m -s /bin/bash -u 2022 -g 2022 ceph-admin && echo ceph-admin:123456 | chpasswd

root@ceph-deploy:~# echo "ceph-admin ALL = NOPASSWD:ALL" | tee /etc/sudoers.d/ceph-admin

ceph-admin ALL = NOPASSWD:ALL

root@ceph-deploy:~# chmod 0440 /etc/sudoers.d/ceph-admin

root@ceph-deploy:~# ll /etc/sudoers.d/ceph-admin

-r--r----- 1 root root 30 Aug 22 03:01 /etc/sudoers.d/ceph-admin

设置免密登录（只需要在ceph-deploy节点上执行）

登录ceph-admin

ceph-admin@ceph-deploy:~$ ssh-keygen -t rsa

ceph-admin@ceph-deploy:~$ ssh-copy-id ceph-admin@ubuntu-ceph1

ceph-admin@ceph-deploy:~$ ssh-copy-id ceph-admin@ubuntu-ceph2

ceph-admin@ceph-deploy:~$ ssh-copy-id ceph-admin@ubuntu-ceph3

3.3部署

安装ceph部署工具

ceph-admin@ceph-deploy:~$ apt-cache madison ceph-deploy

ceph-admin@ceph-deploy:~$ sudo apt install ceph-deploy -y

Ubuntu 各服务器需要单独安装 Python2：

apt install python2.7 -y && ln -sv /usr/bin/python2.7 /usr/bin/python2

初始化mon节点

ceph-admin@ceph-deploy:~$ mkdir ceph-cluster #保存集群的初始化配置信息

ceph-admin@ceph-deploy:~$ cd ceph-cluster/

注意：ceph-deploy部署程序会将文件输出到当前目录。执行ceph-deploy部署时请确保位于ceph-deploy目录中。

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy new --cluster-network 192.168.122.0/24 --public-network 192.168.202.0/24 ubuntu-ceph1 ceph-admin@ceph-deploy:~/ceph-cluster$ ll total 20 drwxrwxr-x 2 ceph-admin ceph-admin 4096 Aug 22 03:30 ./ drwxr-xr-x 4 ceph-admin ceph-admin 4096 Aug 22 03:30 ../ -rw-rw-r-- 1 ceph-admin ceph-admin 274 Aug 22 03:30 ceph.conf -rw-rw-r-- 1 ceph-admin ceph-admin 3404 Aug 22 03:30 ceph-deploy-ceph.log -rw------- 1 ceph-admin ceph-admin 73 Aug 22 03:30 ceph.mon.keyring ceph-admin@ceph-deploy:~/ceph-cluster$ cat ceph.conf [global] fsid = ab782bbc-84bb-4815-8146-914012677b03 public_network = 192.168.202.0/24 cluster_network = 192.168.122.0/24 mon_initial_members = ubuntu-ceph1 mon_host = 192.168.202.142 auth_cluster_required = cephx auth_service_required = cephx auth_client_required = cephx

初始化 node 节点过程

安装Ceph包到指定节点，参数–no-adjust-repos是直接使用本地源，不生成官方源。

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy install --no-adjust-repos --nogpgcheck ubuntu-ceph1 ubuntu-ceph2 ubuntu-ceph3

此过程会在指定的 ceph node 节点按照串行的方式逐个服务器安装 epel 源和 ceph 源并按安装 ceph ceph-radosgw

配置 mon 节点并生成及同步秘钥：

在各 mon 节点按照组件 ceph-mon,并通初始化 mon 节点，mon 节点 ha 还可以后期横向扩容。

安装：

root@ubuntu-ceph1:~# apt install ceph-mon

root@ubuntu-ceph2:~# apt install ceph-mon

root@ubuntu-ceph3:~# apt install ceph-mon

生成密钥：

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy mon create-initial

验证 mon 节点：

验证在 mon 定节点已经自动安装并启动了 ceph-mon 服务，并且后期在 ceph-deploy 节点初始化目录会生成一些 bootstrap ceph mds/mgr/osd/rgw 等服务的 keyring 认证文件，这些初始化文件拥有对 ceph 集群的最高权限，所以一定要保存好。

root@ubuntu-ceph1:~# ps -ef | grep ceph-mon ceph 13606 1 0 05:43 ? 00:00:02 /usr/bin/ceph-mon -f --cluster ceph --id ubuntu-ceph1 --setuser ceph --setgroup ceph root 13910 13340 0 05:53 pts/1 00:00:00 grep --color=auto ceph-mon

分发 admin 秘钥

如果在 ceph-deploy 节点管理集群:

root@ceph-deploy:~# apt install ceph-common -y #先安装 ceph 的公共组件

root@ubuntu-ceph1:~# apt install ceph-common -y

root@ubuntu-ceph2:~# apt install ceph-common -y

root@ubuntu-ceph3:~# apt install ceph-common -y

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy admin ubuntu-ceph1 ubuntu-ceph2 ubuntu-ceph3

节点查看

命令查看

root@ubuntu-ceph1:~# ceph -s cluster: id: c84ff7b7-f8d1-44ac-a87e-e4048e3bebc5 health: HEALTH_WARN mons are allowing insecure global_id reclaim #需要禁用非安全模式通信 services: mon: 3 daemons, quorum ubuntu-ceph1,ubuntu-ceph2,ubuntu-ceph3 (age 22m) mgr: no daemons active osd: 0 osds: 0 up, 0 in data: pools: 0 pools, 0 pgs objects: 0 objects, 0 B usage: 0 B used, 0 B / 0 B avail pgs: root@ubuntu-ceph1:~# ceph config set mon auth_allow_insecure_global_id_reclaim false root@ubuntu-ceph1:~# ceph -s cluster: id: c84ff7b7-f8d1-44ac-a87e-e4048e3bebc5 health: HEALTH_OK services: mon: 3 daemons, quorum ubuntu-ceph1,ubuntu-ceph2,ubuntu-ceph3 (age 24m) mgr: no daemons active osd: 0 osds: 0 up, 0 in data: pools: 0 pools, 0 pgs objects: 0 objects, 0 B usage: 0 B used, 0 B / 0 B avail pgs:

部署 ceph-mgr 节点

初始化 ceph-mgr 节点：

root@ubuntu-ceph1:~# apt install ceph-mgr

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy mgr create ubuntu-ceph1

ceph-deploy 管理 ceph 集群：

在 ceph-deploy 节点配置一下系统环境，以方便后期可以执行 ceph 管理命令。

root@ceph-deploy:~# apt install ceph-common

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy admin ceph-deploy #推送正证书给自己

ceph命令授权给ceph-admin用户

root@ceph-deploy:~# setfacl -m u:ceph-admin:rw /etc/ceph/ceph.client.admin.keyring

准备 OSD 节点

OSD 节点安装运行环境:

#在添加 osd 之前,对node节点安装基本环境:

ceph-admin@ceph-deploy:~/ceph-cluster$ ceph-deploy install --no-adjust-repos --nogpgcheck ubuntu-ceph1 ubuntu-ceph2 ubuntu-ceph3

擦除磁盘:

ceph-deploy disk zap ubuntu-ceph1 /dev/sdb

ceph-deploy disk zap ubuntu-ceph1 /dev/sdc

ceph-deploy disk zap ubuntu-ceph1 /dev/sdd

ceph-deploy disk zap ubuntu-ceph1 /dev/sde

ceph-deploy disk zap ubuntu-ceph2 /dev/sdb

ceph-deploy disk zap ubuntu-ceph2 /dev/sdc

ceph-deploy disk zap ubuntu-ceph2 /dev/sdd

ceph-deploy disk zap ubuntu-ceph2 /dev/sde

ceph-deploy disk zap ubuntu-ceph3 /dev/sdb

ceph-deploy disk zap ubuntu-ceph3 /dev/sdc

ceph-deploy disk zap ubuntu-ceph3 /dev/sdd

ceph-deploy disk zap ubuntu-ceph3 /dev/sde

添加主机的磁盘osd

数据分类保存方式：

Data：即 ceph 保存的对象数据

Block: rocks DB 数据即元数据

block-wal：数据库的 wal 日志

osd的id从0开始顺序使用 0-3

ceph-deploy osd create ubuntu-ceph1 --data /dev/sdb

ceph-deploy osd create ubuntu-ceph1 --data /dev/sdc

ceph-deploy osd create ubuntu-ceph1 --data /dev/sdd

ceph-deploy osd create ubuntu-ceph1 --data /dev/sde

4-7

ceph-deploy osd create ubuntu-ceph2 --data /dev/sdb

ceph-deploy osd create ubuntu-ceph2 --data /dev/sdc

ceph-deploy osd create ubuntu-ceph2 --data /dev/sdd

ceph-deploy osd create ubuntu-ceph2 --data /dev/sde

8-11

ceph-deploy osd create ubuntu-ceph3 --data /dev/sdb

ceph-deploy osd create ubuntu-ceph3 --data /dev/sdc

ceph-deploy osd create ubuntu-ceph3 --data /dev/sdd

ceph-deploy osd create ubuntu-ceph3 --data /dev/sde

测试上传与下载数据

创建pool

存取数据时，客户端必须首先连接至 RADOS 集群上某存储池，然后根据对象名称由相关的

CRUSH 规则完成数据对象寻址。于是，为了测试集群的数据存取功能，这里首先创建一个用

于测试的存储池 mytest，并设定其 PG 数量为 32 个。

ceph-admin@ceph-deploy:~$ ceph osd pool create mytest 32 32 pool \'mytest\' created ceph-admin@ceph-deploy:~$ ceph osd pool ls device_health_metrics mytest ceph-admin@ceph-deploy:~$ ceph pg ls-by-pool mytest | awk \'{print $1,$2,$15}\' #验证 PG 与 PGP 组合 PG OBJECTS ACTING 2.0 0 [7,10,3]p7 2.1 0 [2,8,6]p2 2.2 0 [5,1,9]p5 2.3 0 [5,2,9]p5 2.4 0 [1,10,6]p1 2.5 0 [8,0,4]p8 2.6 0 [1,8,4]p1 2.7 0 [6,10,2]p6 2.8 0 [7,9,0]p7 2.9 0 [1,7,10]p1 2.a 0 [11,3,7]p11 2.b 0 [8,7,2]p8 2.c 0 [11,0,5]p11 2.d 0 [9,4,3]p9 2.e 0 [2,9,7]p2 2.f 0 [8,4,0]p8 2.10 0 [8,1,5]p8 2.11 0 [6,1,9]p6 2.12 0 [10,3,7]p10 2.13 0 [9,4,3]p9 2.14 0 [6,9,0]p6 2.15 0 [9,1,5]p9 2.16 0 [5,11,1]p5 2.17 0 [6,10,2]p6 2.18 0 [9,4,2]p9 2.19 0 [3,6,8]p3 2.1a 0 [6,8,2]p6 2.1b 0 [11,7,3]p11 2.1c 0 [10,7,1]p10 2.1d 0 [10,7,0]p10 2.1e 0 [3,10,5]p3 2.1f 0 [0,7,8]p0 * NOTE: afterwards

上传文件

ceph-admin@ceph-deploy:~$ sudo rados put msg1 /var/log/syslog --pool=mytest # 把 messages 文件上传到 mytest并指定对象 id 为 msg1

列出文件

ceph-admin@ceph-deploy:~$ sudo rados ls --pool=mytest msg1

文件信息

ceph-admin@ceph-deploy:~$ ceph osd map mytest msg1 osdmap e129 pool \'mytest\' (2) object \'msg1\' -> pg 2.c833d430 (2.10) -> up ([8,1,5], p8) acting ([8,1,5], p8)

下载文件

ceph-admin@ceph-deploy:~$ sudo rados get msg1 --pool=mytest /tmp/my.txt