文章目录
部署NFS双机热备高可用环境，用作K8S容器集群的StorageClass，更好地实现K8S数据持久化
思路：
NFS + Keepalived 实现高可用，防止单点故障。
Rsync+Inotify 实现主备间共享数据进行同步。
环境准备
两台NFS服务器
192.168.20.130
192.168.20.131
一个VIP：192.168.20.251
一个k8s集群
通过在k8s部署nfs-client-provisioner连接NFS服务器的VIP使用NAS存储
技术要求
- 两个NFS节点机器的配置要一致
- keepalived监控nfs进程，master的nfs主进程宕掉无法启动时由slave的nfs接管继续工作。
- k8s数据备份到slave，同时master和slave数据用rsync+inotify实时同步，保证数据完整性。
- 生产环境下，最好给NFS共享目录单独挂载一块硬盘或单独的磁盘分区。
安装NFS服务
1）安装nfs
# yum -y install nfs-utils
2）创建nfs共享目录
# mkdir /home/k8s_storage
3）编辑export文件
# vim /etc/exports
/home/k8s_storage 192.168.20.0/24(rw,sync,no_root_squash)
4）配置生效
# exportfs -r
5）查看生效
# exportfs
6）启动rpcbind、nfs服务
# systemctl restart rpcbind && systemctl enable rpcbind
# systemctl restart nfs && systemctl enable nfs
7）查看 RPC 服务的注册状况
# rpcinfo -p localhost
8）showmount测试
Master节点测试
# showmount -e 192.168.20.130
Export list for 192.168.20.130:
/home/k8s_storage 192.168.20.0/24
Slave节点测试
# showmount -e 192.168.20.131
Export list for 192.168.20.131:
/home/k8s_storage 192.168.20.0/24
安装keepalived
1）安装keepalived
# yum -y install keepalived
2）Master节点的keepalived.conf配置
这里特别需要注意：
一定要设置keepalived为非抢占模式，如果设置成抢占模式会在不断的切换主备时容易造成NFS数据丢失。
# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf_bak
# >/etc/keepalived/keepalived.conf
# vi /etc/keepalived/keepalived.conf
global_defs {
router_id master #id可以随便设
}
vrrp_script chk_nfs {
script "/etc/keepalived/nfs_check.sh" #监控脚本
interval 2
weight -20 #keepalived部署了两台，所以设为20，如果三台就设为30
}
vrrp_instance VI_1 {
state BACKUP #两台主机都设为backup非抢占模式
interface ens192 #网卡名写自己的，不要照抄
virtual_router_id 51
priority 100 #服务启动优先级，值越大，优先级越高，BACKUP 不能大于MASTER
advert_int 1
nopreempt #设置为非抢占模式必须要该参数
authentication {
auth_type PASS
auth_pass Kibo
}
track_script {
chk_nfs
}
virtual_ipaddress {
192.168.20.251 #虚拟ip
}
}
3）Slave节点的keepalived.conf配置
只需将priority参数项修改为80，其他配置的和master节点一样，脚本也一样。
# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf_bak
# >/etc/keepalived/keepalived.conf
# vi /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
router_id master
}
vrrp_script chk_nfs {
script "/etc/keepalived/nfs_check.sh"
interval 2
weight -20
}
vrrp_instance VI_1 {
state BACKUP
interface eth0
virtual_router_id 51
priority 80
advert_int 1
nopreempt
authentication {
auth_type PASS
auth_pass 1111
}
track_script {
chk_nfs
}
virtual_ipaddress {
192.168.20.251
}
}
4）编辑nfs_check.sh监控脚本
# vi /etc/keepalived/nfs_check.sh
#!/bin/bash
A=`ps -C nfsd --no-header | wc -l`
if [ $A -eq 0 ];then
systemctl restart nfs-server.service
sleep 2
if [ `ps -C nfsd --no-header| wc -l` -eq 0 ];then
pkill keepalived
fi
fi
设置脚本执行权限
# chmod 755 /etc/keepalived/nfs_check.sh
5）启动keepalived服务
# systemctl restart keepalived.service && systemctl enable keepalived.service
查看服务进程是否启动
# ps -ef|grep keepalived
6）检查vip是否存在
在两台节点机器上执行"ip addr"命令查看vip，其中会在一台机器上产生vip地址。
# ip addr|grep 192.168.20.251
inet 192.168.20.251/32 scope global ens192
测试vip地址要能被ping通
# ping 192.168.20.52
PING 192.168.20.52 (192.168.20.52) 56(84) bytes of data.
64 bytes from 192.168.20.52: icmp_seq=1 ttl=64 time=0.068 ms
64 bytes from 192.168.20.52: icmp_seq=2 ttl=64 time=0.027 ms
64 bytes from 192.168.20.52: icmp_seq=3 ttl=64 time=0.041 ms
7）keepalived故障测试
停掉vip所在的Master节点机器上的keepalived服务后，发现vip会自动飘移到另一台Backup机器上才算测试成功。
当该Master节点的keepalived服务重新启动后，vip不会重新飘移回来。因为keepalived采用了非抢占模式。
如果keepalived设置为抢占模式，vip会在Master节点的keepalived重启恢复后自动飘回去，
但是这样一直来回切换可能会造成NFS数据不完整，因为这里必须设置成非抢占模式。
由于配置了nfs的nfs_check.sh监控脚本，所以当其中一台节点机器上的NFS服务宕停后会自动重启NFS。
如果NFS服务重启失败，则会自动关闭该节点机器上的keepalived服务，如果该节点有vip则会自动飘移到另外一台节点上。
安装Rsync+Inofity
1）安装rsync和inotify
# yum -y install rsync inotify-tools
2）Master节点机器配置rsyncd.conf
# cp /etc/rsyncd.conf /etc/rsyncd.conf_bak
# >/etc/rsyncd.conf
# vi /etc/rsyncd.conf
uid = root
gid = root
use chroot = 0
port = 873
hosts allow = 192.168.20.0/24 #允许ip访问设置，可以指定ip或ip段
max connections = 0
timeout = 300
pid file = /var/run/rsyncd.pid
lock file = /var/run/rsyncd.lock
log file = /var/log/rsyncd.log
log format = %t %a %m %f %b
transfer logging = yes
syslog facility = local3
[master_web]
path = /home/k8s_storage
comment = master_web
ignore errors
read only = no #是否允许客户端上传文件
list = no
auth users = rsync #指定由空格或逗号分隔的用户名列表，只有这些用户才允许连接该模块
secrets file = /etc/rsyncd.passwd #保存密码和用户名文件，需要自己生成
编辑密码和用户文件（格式为"用户名:密码"）
# vim /etc/rsyncd.passwd
rsync:Kibo@2020
编辑同步密码（注意这个文件和上面的密码和用户文件路径不一样）
该文件内容只需要填写从服务器的密码，例如这里从服务器配的用户名密码都是rsync:123456，则主服务器则写123456一个就可以了
# vim /opt/rsyncd.passwd
Kibo@2020
设置文件执行权限
# chmod 600 /etc/rsyncd.passwd && chmod 600 /opt/rsyncd.passwd
启动服务
# systemctl enable rsyncd && systemctl restart rsyncd
检查rsync服务进程是否启动
# ps -ef|grep rsync
3）Slave节点机器配置rsyncd.conf
就把master主机/etc/rsyncd.conf配置文件里的[master_web]改成[slave_web]
其他都一样，密码文件也设为一样
# cp /etc/rsyncd.conf /etc/rsyncd.conf_bak
# >/etc/rsyncd.conf
# vim /etc/rsyncd.conf
uid = root
gid = root
use chroot = 0
port = 873
hosts allow = 192.168.20.0/24
max connections = 0
timeout = 300
pid file = /var/run/rsyncd.pid
lock file = /var/run/rsyncd.lock
log file = /var/log/rsyncd.log
log format = %t %a %m %f %b
transfer logging = yes
syslog facility = local3
[slave_web]
path = /home/k8s_storage
comment = master_web
ignore errors
read only = no
list = no
auth users = rsync
secrets file = /etc/rsyncd.passwd
编辑密码和用户文件（格式为"用户名:密码"）
# vim /etc/rsyncd.passwd
rsync:123456
编辑同步密码
# vim /opt/rsyncd.passwd
123456
设置文件执行权限
# chmod 600 /etc/rsyncd.passwd
# chmod 600 /opt/rsyncd.passwd
启动服务
# systemctl enable rsyncd && systemctl restart rsyncd
检查rsync服务进程是否启动
# ps -ef|grep rsync
手动验证下Master节点NFS数据同步到Slave节点
在Master节点的NFS共享目录下创建测试数据
# ls /home/k8s_storage/
# mkdir /home/k8s_storage/test
# touch /home/k8s_storage/{a,b}
# ls /home/k8s_storage/
a b test
手动同步Master节点的NFS共享目录数据到Slave节点的NFS共享目录下
# rsync -avzp --delete /home/k8s_storage/ rsync@192.168.20.131::slave_web --password-file=/opt/rsyncd.passwd
到Slave节点查看
# ls /home/k8s_storage/
a b test
上面rsync同步命令说明：
- /home/k8s_storage/ 是同步的NFS共享目录
- rsync@192.168.20.131::slave_web
- rsync 是Slave节点服务器的/etc/rsyncd.passwd文件中配置的用户名
- 192.168.20.131为Slave节点服务ip
- slave_web 为Slave服务器的rsyncd.conf中配置的同步模块名
- –password-file=/opt/rsyncd.passwd 是Master节点同步到Slave节点使用的密码文件，文件中配置的是Slave节点服务器的/etc/rsyncd.passwd文件中配置的密码
设置Rsync+Inotify自动同步
这里需要注意：不能设置Master和Slave节点同时执行rsync自动同步，即不能同时设置双向同步。因为Master节点将数据同步到Slave节点，如果Slave节点再将数据同步回到Master节点，这个就矛盾了。所以需要确保只有一方在执行自动同步到另一方的操作。方式就是判断当前节点服务器是否存在VIP，如存在VIP则自动同步数据到另一台节点上。如不存在VIP则不执行自动同步操作。
Master节点服务器操作
编写自动同步脚本/opt/rsync_inotify.sh
#!/bin/bash
host=192.168.20.131
src=/home/k8s_storage/
des=slave_web
password=/opt/rsyncd.passwd
user=rsync
inotifywait=/usr/bin/inotifywait
$inotifywait -mrq --timefmt '%Y%m%d %H:%M' --format '%T %w%f%e' -e modify,delete,create,attrib $src \
| while read files ;do
rsync -avzP --delete --timeout=100 --password-file=${password} $src $user@$host::$des
echo "${files} was rsynced" >>/tmp/rsync.log 2>&1
done
编写VIP监控脚本/opt/vip_monitor.sh
#!/bin/bash
VIP_NUM=`ip addr|grep 192.168.20.251|wc -l`
RSYNC_INOTIRY_NUM=`ps -ef|grep /usr/bin/inotifywait|grep -v grep|wc -l`
if [ ${VIP_NUM} -ne 0 ];then
echo "VIP在当前NFS节点服务器上" >/dev/null 2>&1
if [ ${RSYNC_INOTIRY_NUM} -ne 0 ];then
echo "rsync_inotify.sh脚本已经在后台执行中" >/dev/null 2>&1
else
echo "需要在后台执行rsync_inotify.sh脚本" >/dev/null 2>&1
nohup sh /opt/rsync_inotify.sh &
fi
else
echo "VIP不在当前NFS节点服务器上" >/dev/null 2>&1
if [ ${RSYNC_INOTIRY_NUM} -ne 0 ];then
echo "需要关闭后台执行的rsync_inotify.sh脚本" >/dev/null 2>&1
ps -ef|grep rsync_inotify.sh|grep -v grep|awk '{print $2}'|xargs kill -9
ps -ef|grep inotifywait|grep -v grep|awk '{print $2}'|xargs kill -9
else
echo "rsync_inotify.sh脚本当前未执行" >/dev/null 2>&1
fi
fi
编写持续执行脚本/opt/rsync_monit.sh
#!/bin/bash
while [ "1" = "1" ]
do
/bin/bash -x /opt/vip_monitor.sh >/dev/null 2>&1
done
后台运行脚本
# chmod 755 /opt/rsync_inotify.sh
# chmod 755 /opt/vip_monitor.sh
# chmod 755 /opt/rsync_monit.sh
# nohup sh /opt/rsync_inotify.sh &
# nohup sh /opt/rsync_monit.sh &
设置rsync_monit.sh脚本的开机启动
# chmod +x /etc/rc.d/rc.local
# echo "nohup sh /opt/rsync_monit.sh & " >> /etc/rc.d/rc.local
Slave节点服务器操作
脚本名为/opt/rsync_inotify.sh，内容如下：
#!/bin/bash
host=192.168.20.130
src=/home/k8s_storage/
des=master_web
password=/opt/rsyncd.passwd
user=rsync
inotifywait=/usr/bin/inotifywait
$inotifywait -mrq --timefmt '%Y%m%d %H:%M' --format '%T %w%f%e' -e modify,delete,create,attrib $src \
| while read files ;do
rsync -avzP --delete --timeout=100 --password-file=${password} $src $user@$host::$des
echo "${files} was rsynced" >>/tmp/rsync.log 2>&1
done
编写VIP监控脚本/opt/vip_monitor.sh
#!/bin/bash
VIP_NUM=`ip addr|grep 192.168.20.251|wc -l`
RSYNC_INOTIRY_NUM=`ps -ef|grep /usr/bin/inotifywait|grep -v grep|wc -l`
if [ ${VIP_NUM} -ne 0 ];then
echo "VIP在当前NFS节点服务器上" >/dev/null 2>&1
if [ ${RSYNC_INOTIRY_NUM} -ne 0 ];then
echo "rsync_inotify.sh脚本已经在后台执行中" >/dev/null 2>&1
else
echo "需要在后台执行rsync_inotify.sh脚本" >/dev/null 2>&1
nohup sh /opt/rsync_inotify.sh &
fi
else
echo "VIP不在当前NFS节点服务器上" >/dev/null 2>&1
if [ ${RSYNC_INOTIRY_NUM} -ne 0 ];then
echo "需要关闭后台执行的rsync_inotify.sh脚本" >/dev/null 2>&1
ps -ef|grep rsync_inotify.sh|grep -v grep|awk '{print $2}'|xargs kill -9
ps -ef|grep inotifywait|grep -v grep|awk '{print $2}'|xargs kill -9
else
echo "rsync_inotify.sh脚本当前未执行" >/dev/null 2>&1
fi
fi
编写持续执行脚本/opt/rsync_monit.sh
#!/bin/bash
while [ "1" = "1" ]
do
/bin/bash -x /opt/vip_monitor.sh >/dev/null 2>&1
done
后台运行脚本 (只执行rsync_monit.sh)
# chmod 755 /opt/rsync_inotify.sh
# chmod 755 /opt/vip_monitor.sh
# chmod 755 /opt/rsync_monit.sh
# nohup sh /opt/rsync_monit.sh &
设置rsync_monit.sh脚本的开机启动
# chmod +x /etc/rc.d/rc.local
# echo "nohup sh /opt/rsync_monit.sh & " >> /etc/rc.d/rc.local
验证自动同步
1）比如当前VIP在Master节点，在Master节点创建测试数据，观察是否自动同步到Slave节点
# ip addr|grep 192.168.20.251
inet 192.168.20.251/32 scope global ens192
# rm -rf /home/k8s_storage/*
# echo "test" > /home/k8s_storage/haha
# ls /home/k8s_storage/
haha
到Slave节点上查看，已自动同步过来
# ls /home/k8s_storage/
haha
# cat /home/k8s_storage/haha
test
2）接着关闭Master节点的keeplived，将VIP飘移到Slave节点
# systemctl stop keepalived
# ip addr|grep 192.168.20.251
到Slave节点上查看，发现VIP已经飘移过来了
# ip addr|grep 192.168.20.251
inet 192.168.20.251/32 scope global eth0
在Slave节点创建测试数据，观察是否自动同步到Master节点
# rm -rf /home/k8s_storage/*
# mkdir /home/k8s_storage/cha
# echo "heihei" > /home/k8s_storage/you
到Master节点查看，发现数据已经同步过来了
# ls /home/k8s_storage/
cha heihei
3）模拟Master节点和Slave节点关机，观察开机后：
/opt/rsync_monit.sh脚本会实现开机自启动。
按照上面Master和Slave节点的自动同步验证OK。
K8s部署NFS-Provisioner
要使用 StorageClass，我们就得安装对应的自动配置程序，比如我们这里存储后端使用的是 nfs，那么我们就需要使用到一个 nfs-client 的自动配置程序，我们也叫它 Provisioner，这个程序使用我们已经配置好的 nfs 服务器，来自动创建持久卷，也就是自动帮我们创建 PV。前提k8s集群每台服务器安装nfs客户端【yum install -y nfs-utils】。
创建一个ServiceAccount
新建的一个名为 nfs-client-provisioner 的ServiceAccount，然后绑定了一个名为 nfs-client-provisioner-runner 的ClusterRole，而该ClusterRole声明了一些权限，其中就包括对persistentvolumes的增、删、改、查等权限，所以我们可以利用该ServiceAccount来自动创建 PV
cat nfs-rbac.yaml
apiVersion: v1
kind: ServiceAccount
metadata:
name: nfs-client-provisioner
# replace with namespace where provisioner is deployed
namespace: k8s-nfs
---
kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: nfs-client-provisioner-runner
rules:
- apiGroups: [""]
resources: ["persistentvolumes"]
verbs: ["get", "list", "watch", "create", "delete"]
- apiGroups: [""]
resources: ["persistentvolumeclaims"]
verbs: ["get", "list", "watch", "update"]
- apiGroups: ["storage.k8s.io"]
resources: ["storageclasses"]
verbs: ["get", "list", "watch"]
- apiGroups: [""]
resources: ["events"]
verbs: ["list", "watch", "create", "update", "patch"]
- apiGroups: [""]
resources: ["endpoints"]
verbs: ["create", "delete", "get", "list", "watch", "patch", "update"]
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: run-nfs-client-provisioner
subjects:
- kind: ServiceAccount
name: nfs-client-provisioner
# replace with namespace where provisioner is deployed
namespace: k8s-nfs
roleRef:
kind: ClusterRole
name: nfs-client-provisioner-runner
apiGroup: rbac.authorization.k8s.io
---
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: leader-locking-nfs-client-provisioner
# replace with namespace where provisioner is deployed
namespace: k8s-nfs
rules:
- apiGroups: [""]
resources: ["endpoints"]
verbs: ["get", "list", "watch", "create", "update", "patch"]
---
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: leader-locking-nfs-client-provisioner
# replace with namespace where provisioner is deployed
namespace: k8s-nfs
subjects:
- kind: ServiceAccount
name: nfs-client-provisioner
# replace with namespace where provisioner is deployed
namespace: k8s-nfs
roleRef:
kind: Role
name: leader-locking-nfs-client-provisioner
apiGroup: rbac.authorization.k8s.io
部署 NFS Provisioner
cat nfs-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: nfs-client-provisioner
labels:
app: nfs-client-provisioner
# replace with namespace where provisioner is deployed
namespace: k8s-nfs
spec:
replicas: 1
strategy:
type: Recreate
selector:
matchLabels:
app: nfs-client-provisioner
template:
metadata:
labels:
app: nfs-client-provisioner
spec:
serviceAccountName: nfs-client-provisioner
containers:
- name: nfs-client-provisioner
#image: quay.io/external_storage/nfs-client-provisioner:latest
image: docker.io/jmgao1983/nfs-client-provisioner:latest
volumeMounts:
- name: nfs-client-root
mountPath: /persistentvolumes
env:
- name: PROVISIONER_NAME #nfs-provisioner的名称
value: fuseim.pri/ifs
- name: NFS_SERVER #nfs服务端地址
value: 192.168.20.251
- name: NFS_PATH #nfs服务端路径
value: /home/k8s_storage
volumes:
- name: nfs-client-root
nfs:
server: 192.168.20.251 #nfs服务端地址
path: /home/k8s_storage #nfs服务端路径
设置默认存储
cat nfs-storage.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: managed-nfs-storage
annotations:
storageclass.kubernetes.io/is-default-class: "true" #---设置为默认的storageclass
provisioner: fuseim.pri/ifs # or choose another name, must match deployment's env PROVISIONER_NAME'
parameters:
archiveOnDelete: "false"
创建 NFS StorageClass
cat nfs-storage.yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: managed-nfs-storage
annotations:
storageclass.kubernetes.io/is-default-class: "true" #---设置为默认的storageclass
provisioner: fuseim.pri/ifs # or choose another name, must match deployment's env PROVISIONER_NAME'
parameters:
archiveOnDelete: "false"
上面把StorageClass资源对象创建成功了,可以通过创建PVC使用了