By fanux | May 29, 2018
广告 | kubernetes各版本离线安装包
1.11.1版本已经修复此bug，如在centos下使用ipvs模式建议升级到v1.11.1,包地址
1.11.0版本ipset bug说明
1.11.0版本centos下使用ipvs模式会出问题 65461
Jun 25 20:50:00 VM_3_4_centos kube-proxy[3828]: E0625 20:50:00.312569 3828 ipset.go:156] Failed to make sure ip set: &{{KUBE-LOOP-BACK hash:ip,port,ip inet 1024 65536 0-65535 Kubernetes endpoints dst ip:port, source ip for solving hairpin purpose} map[] 0xc42073e1d0} exist, error: error creating ipset KUBE-LOOP-BACK, error: exit status 2
主要是ipset不支持comment:
[root@compute063 ~]# ipset create foo hash:ip comment
ipset v6.19: Unknown argument: `comment'
Try `ipset help' for more information.
尝试升级ipset问题依然没解决
[root@izrj9auny05eigffvcosvbz ipset-6.38]# ipset create foo hash:ip comment
ipset v6.38: Argument `comment' is supported in the kernel module of the set type hash:ip starting from the revision 2 and you have installed revision 1 only. Your kernel is behind your ipset utility.
Try `ipset help' for more information.
在不改kubernetes情况下可以通过升级内核和ipset解决
升级内核
rpm -ivh kernel-4.14.49-1.x86_64.rpm
rpm -ivh kernel-devel-4.14.49-1.x86_64.rpm
修改grub配置，默认启动新内核
vi /etc/default/grub
修改成 GRUB_DEFAULT=0
grub2-mkconfig -o /boot/grub2/grub.cfg
ipset 安装过程
yum install -y kernel-devel
yum install -y bzip2
wget http://ipset.netfilter.org/ipset-6.38.tar.bz2
cd ipset-6.38
bzip2 -d ipset-6.38.tar.bz2
tar xvf ipset-6.38.tar
cd /lib/modules/3.10.0-693.2.2.el7.x86_64
ln -s /usr/src/kernels/3.10.0-862.3.3.el7.x86_64 build
./configure && make && make install
kubernetes启用ipvs
确保内核开启了ipvs模块
[root@k8s ~]# lsmod|grep ip_vs
ip_vs_sh 12688 0
ip_vs_wrr 12697 0
ip_vs_rr 12600 16
ip_vs 141092 23 ip_vs_rr,ip_vs_sh,xt_ipvs,ip_vs_wrr
nf_conntrack 133387 9 ip_vs,nf_nat,nf_nat_ipv4,nf_nat_ipv6,xt_conntrack,nf_nat_masquerade_ipv4,nf_conntrack_netlink,nf_conntrack_ipv4,nf_conntrack_ipv6
libcrc32c 12644 3 ip_vs,nf_nat,nf_conntrack
没开启加载方式:
modprobe -- ip_vs
modprobe -- ip_vs_rr
modprobe -- ip_vs_wrr
modprobe -- ip_vs_sh
modprobe -- nf_conntrack_ipv4
1.10以上版本,使用kubeadm安装的，直接修改kube-proxy configmap即可
kubectl edit configmap kube-proxy -n kube-system
ipvs:
minSyncPeriod: 0s
scheduler: ""
syncPeriod: 30s
kind: KubeProxyConfiguration
metricsBindAddress: 127.0.0.1:10249
mode: "ipvs" # 加上这个
nodePortAddresses: null
看到pod如下信息表明成功
[root@k8s ~]# kubectl logs kube-proxy-72lg9 -n kube-system
I0530 03:38:11.455609 1 feature_gate.go:226] feature gates: &{{} map[]}
I0530 03:38:11.490470 1 server_others.go:183] Using ipvs Proxier.
W0530 03:38:11.503868 1 proxier.go:304] IPVS scheduler not specified, use rr by default
I0530 03:38:11.504109 1 server_others.go:209] Tearing down inactive rules.
I0530 03:38:11.552587 1 server.go:444] Version: v1.10.3
安装ipvsadm工具
yum install -y ipvsadm
检查service ipvs配置：
[root@k8s ~]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 172.31.244.239:32000 rr
-> 192.168.77.9:8443 Masq 1 0 0
TCP 172.31.244.239:32001 rr
-> 192.168.77.8:3000 Masq 1 0 0
TCP 10.96.0.1:443 rr persistent 10800
-> 172.31.244.239:6443 Masq 1 0 0
TCP 10.96.0.10:53 rr
-> 192.168.77.7:53 Masq 1 0 0
-> 192.168.77.10:53 Masq 1 0 0
TCP 10.96.82.0:80 rr
-> 192.168.77.8:3000 Masq 1 0 0
TCP 10.96.152.25:8086 rr
-> 192.168.77.12:8086 Masq 1 0 0
TCP 10.96.232.136:6666 rr
可以看到我们的dashboard dns什么的都已经配置了，可以验证一下：
[root@k8s ~]# wget https://172.31.244.239:32000 --no-check-certificate
--2018-05-30 16:17:15-- https://172.31.244.239:32000/
正在连接 172.31.244.239:32000... 已连接。
警告: 无法验证 172.31.244.239 的由 “/CN=.” 颁发的证书:
出现了自己签名的证书。
警告: 证书通用名 “.” 与所要求的主机名 “172.31.244.239” 不符。
已发出 HTTP 请求，正在等待回应... 200 OK
长度：990 [text/html]
正在保存至: “index.html”
100%[=======================================================================================================================================================>] 990 --.-K/s 用时 0s
2018-05-30 16:17:15 (16.3 MB/s) - 已保存 “index.html” [990/990])
是通的，完全ok
这里十分推荐大家使用ipvs模式，iptables出了问题不好调试，而且规则一多性能显著下降，我们甚至出现规则丢失的情况，ipvs稳定很多。
comments powered by Disqus