kolla-ansible 部署OpenStack queens版本笔记

最新推荐文章于 2024-06-20 08:34:37 发布

Dolphinsz

最新推荐文章于 2024-06-20 08:34:37 发布

阅读量8k

点赞数

分类专栏： Linux OpenStack CentOS7

本文链接：https://blog.csdn.net/dolphinsz/article/details/83049521

版权

Linux 同时被 3 个专栏收录

13 篇文章 0 订阅

订阅专栏

CentOS7

10 篇文章 0 订阅

订阅专栏

OpenStack

4 篇文章 0 订阅

订阅专栏

kolla-ansible 部署OpenStack queens版本笔记

一. 实验环境：

6台主机安装CentOS7 minimal系统32G内存，1T+500G双硬盘（其中一个为后期ceph部署做准备），三张千兆网卡（一张用于后期ceph（enp3s0f0），另两张网卡分别作为控制网络（enp0s31f6）以及neutron桥接（enp3s0f1）网络）
网络规划：

host	IP address	remark
controller01	10.132.226.51	1
controller02	10.132.226.52	2
controller03	10.132.226.53	3
compute01	10.132.226.54	4
compute02	10.132.226.55	5
kolla	10.132.226.200	6
virtulal IP	10.132.226.70
虚拟地址池	10.132.226.71-99

二. 控制以及计算节点初始化操作：

使用以下脚本对每个计算机进行初始化配置（kolla除外）执行 sh initnode.sh n(n代表第几台主机)

# /usr/bin/bash

ls -l /etc/sysconfig/network-scripts|awk '/ifcfg-enp[0-9]*/ {print $9}' > default_name.txt
i=5
c=0
cat default_name.txt | while read line
do
    cd /etc/sysconfig/network-scripts
    cp $line ${line}.bak
    sed -i 's/BOOTPROTO=dhcp/BOOTPROTO=static/g' $line
    sed -i 's/ONBOOT=no/ONBOOT=yes/g' $line
    c=$(expr $c + 1)
    if [ "${c}" == "1" ]; then
        continue;
    elif [ "${c}" == "2" ]; then
        i=$(expr $i + 2)
        echo -e "\nIPADDR=10.132.226.${i}${1}" >> $line
        echo "NETMASK=255.255.255.0" >> $line
        echo "GATEWAY=10.132.226.254" >> $line
        echo "DNS1=192.168.0.1" >> $line
        echo "DNS2=114.114.114.114" >> $line
    else
        sed -i '1,11d' $line
        sed -i '1i\OVS_BRIDGE=br-ex' $line
        sed -i '1i\DEVICETYPE=ovs' $line
        sed -i '1i\BOOTPROTO=none' $line
        sed -i '1i\TYPE=OVSPort' $line
    fi
done

systemctl stop firewalld && systemctl disable firewalld && systemctl status firewalld
yum update -y
yum install -y wget vim net-tools
wget -P /etc/yum.repos.d/ https://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
yum install -y docker-ce
mkdir -pv /etc/docker
systemctl restart docker && systemctl status docker
echo -e "\n10.132.226.200\tkolla\n10.132.226.51\tcontroller01\n10.132.226.52\tcontroller02\n10.132.226.53\tcontroller03\n10.132.226.54\tcompute01\n10.132.226.55\tcompute02" >> /etc/hosts
if [ $1 -lt 4 ]; then
	echo "controller0${1}" > /etc/hostname
else
	num=$(expr $1 - 3)
	echo "compute0${num}" > /etc/hostname
fi
reboot

各节点主机初始化内容：
- 配置网卡信息
- 关闭防火墙
- 安装docker
- 修改hostname以及添加hosts信息

设置各节点主机之间免密登录

在各主机依次执行

ssh-keygen
ssh-copy-id -i ~/.ssh/id_rsa.pub root@controller01

将authorized_key文件发放到各主机的~/.ssh/目录

scp authorized_keys kolla:~/.ssh/
scp authorized_keys controller01:~/.ssh/
scp authorized_keys controller02:~/.ssh/
scp authorized_keys controller03:~/.ssh/
scp authorized_keys compute01:~/.ssh/
scp authorized_keys compute02:~/.ssh/

三. kolla主机配置

安装docker：

从阿里云下载docker的repo文件：

[root@kolla ~]# wget -P /etc/yum.repos.d/ https://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo

安装docker-ce：

[root@kolla ~]# yum install -y docker-ce

配置国内镜像：

[root@kolla ~]# mkdir -p /etc/docker
[root@kolla ~]# vim /etc/docker/daemon.json
{
    "registry-mirrors": ["https://7g5a4z30.mirror.aliyuncs.com"]
}

启动docker

[root@kolla ~]# systemctl daemon-reload && systemctl enable docker && systemctl restart docker

检查镜像站点配置是否正确

[root@kolla ~]# docker pull hello-world

安装依赖软件

安装pip并更新

[root@kolla ~]# yum insatll epel-release -y
[root@kolla ~]# yum insatll python-pip -y
[root@kolla ~]# pip install -U pip

修改pip源

[root@kolla ~]# mkdir ~/.pip
[root@kolla ~]# vim ~/.pip/pip.conf
[global]
trusted-host = pypi.douban.com
index-url = http://pypi.douban.com/simple

安装其他依赖包

[root@kolla ~]# yum install python-devel libffi-devel gcc openssl-devel libselinux-python -y

安装配置ansible：
- 先使用pip安装再使用yum安装，可以防止某些py包版本太低
```
[root@kolla ~]# pip install ansible
[root@kolla ~]# yum install ansible -y
```
- 在/etc/ansible/ansible.cfg配置文件中添加以下内容：
```
[defaults]
host_key_checking=False
pipelining=True
forks=100
```
安装配置kolla-ansible：
- 使用pip安装kolla-ansible：
```
pip install kolla-ansible
```
- 复制global.yml和password.yml文件到/etc/kolla目录:
```
cp -r /usr/share/kolla-ansible/etc_examples/kolla /etc/kolla/
```
- 复制all-in-one 和multinode 文件到当前操作目录：
```
cp /usr/share/kolla-ansible/ansible/inventory/* .
```
- 修改global.yml文件
  global.yml
- 拉取镜像
```
kolla-ansible pull -vvv
```
- 再次修改global.yml文件（因为上一个文件拉取的镜像缺少nova-compute等镜像）
  global.yml
- 拉取镜像
```
kolla-ansible pull -vvv
```

上传镜像到本地registry仓库：

配置Docker共享挂载：

[root@kolla ~]# mkdir -p /etc/systemd/system/docker.service.d
[root@kolla ~]# vim /etc/systemd/system/docker.service.d/kolla.conf
[Service]
MountFlags=shared
[root@kolla ~]# systemctl daemon-reload && systemctl restart docker && systemctl status docker

启动registry容器，并将端口映射到4000端口

[root@kolla /]# docker run -d --name registry --restart=always -p 4000:5000 -v /opt/registry:/var/lib/registry registry:2.6.2

修改Docker服务配置，信任本地Registry服务

[root@kolla /]# vim /usr/lib/systemd/system/docker.service
ExecStart=/usr/bin/dockerd --insecure-registry kolla:4000

重新启动docker服务

systemctl daemon-reload && systemctl restart docker

测试registry服务是否正常：

[root@kolla ~]# curl -X GET http://kolla:4000/v2/_catalog
{"repositories":[]}

修改镜像tag：

for i in `docker images|grep -v registry|grep -v R|awk '{print $1}'`;do docker image tag $i:queens kolla:4000/$i:queens;done

push到本地库

for i in `docker images|grep kolla:4000|awk '{print $1}'`;do docker push $i:queens;done

查看镜像是否上传成功：

curl -XGET http://kolla:4000/v2/_catalog
{
	"repositories": [
	"kolla/centos-source-aodh-api",
	"kolla/centos-source-aodh-evaluator",
	"kolla/centos-source-aodh-listener",
	"kolla/centos-source-aodh-notifier",
	"kolla/centos-source-barbican-api",
	"kolla/centos-source-barbican-keystone-listener",
	"kolla/centos-source-barbican-worker",
	"kolla/centos-source-blazar-api",
	"kolla/centos-source-blazar-manager",
	"kolla/centos-source-ceilometer-central",
	"kolla/centos-source-ceilometer-compute",
	"kolla/centos-source-ceilometer-notification",
	"kolla/centos-source-ceph-mds",
	"kolla/centos-source-ceph-mgr",
	"kolla/centos-source-ceph-mon",
	"kolla/centos-source-ceph-nfs",
	"kolla/centos-source-ceph-osd",
	"kolla/centos-source-ceph-rgw",
	"kolla/centos-source-chrony",
	"kolla/centos-source-cinder-api",
	"kolla/centos-source-cinder-backup",
	"kolla/centos-source-cinder-scheduler",
	"kolla/centos-source-cinder-volume",
	"kolla/centos-source-cloudkitty-api",
	"kolla/centos-source-cloudkitty-processor",
	"kolla/centos-source-collectd",
	"kolla/centos-source-congress-api",
	"kolla/centos-source-congress-datasource",
	"kolla/centos-source-congress-policy-engine",
	"kolla/centos-source-cron",
	"kolla/centos-source-designate-api",
	"kolla/centos-source-designate-backend-bind9",
	"kolla/centos-source-designate-central",
	"kolla/centos-source-designate-mdns",
	"kolla/centos-source-designate-producer",
	"kolla/centos-source-designate-sink",
	"kolla/centos-source-designate-worker",
	"kolla/centos-source-dnsmasq",
	"kolla/centos-source-elasticsearch",
	"kolla/centos-source-etcd",
	"kolla/centos-source-fluentd",
	"kolla/centos-source-freezer-api",
	"kolla/centos-source-glance-api",
	"kolla/centos-source-gnocchi-api",
	"kolla/centos-source-gnocchi-metricd",
	"kolla/centos-source-gnocchi-statsd",
	"kolla/centos-source-grafana",
	"kolla/centos-source-haproxy",
	"kolla/centos-source-heat-api",
	"kolla/centos-source-heat-api-cfn",
	"kolla/centos-source-heat-engine",
	"kolla/centos-source-horizon",
	"kolla/centos-source-influxdb",
	"kolla/centos-source-ironic-api",
	"kolla/centos-source-ironic-conductor",
	"kolla/centos-source-ironic-inspector",
	"kolla/centos-source-ironic-pxe",
	"kolla/centos-source-iscsid",
	"kolla/centos-source-karbor-api",
	"kolla/centos-source-karbor-operationengine",
	"kolla/centos-source-karbor-protection",
	"kolla/centos-source-keepalived",
	"kolla/centos-source-keystone",
	"kolla/centos-source-kibana",
	"kolla/centos-source-kolla-toolbox",
	"kolla/centos-source-kuryr-libnetwork",
	"kolla/centos-source-magnum-api",
	"kolla/centos-source-magnum-conductor",
	"kolla/centos-source-manila-api",
	"kolla/centos-source-manila-data",
	"kolla/centos-source-manila-scheduler",
	"kolla/centos-source-manila-share",
	"kolla/centos-source-mariadb",
	"kolla/centos-source-memcached",
	"kolla/centos-source-mistral-api",
	"kolla/centos-source-mistral-engine",
	"kolla/centos-source-mistral-executor",
	"kolla/centos-source-mongodb",
	"kolla/centos-source-multipathd",
	"kolla/centos-source-murano-api",
	"kolla/centos-source-murano-engine",
	"kolla/centos-source-neutron-bgp-dragent",
	"kolla/centos-source-neutron-dhcp-agent",
	"kolla/centos-source-neutron-l3-agent",
	"kolla/centos-source-neutron-lbaas-agent",
	"kolla/centos-source-neutron-metadata-agent",
	"kolla/centos-source-neutron-openvswitch-agent",
	"kolla/centos-source-neutron-server",
	"kolla/centos-source-neutron-server-opendaylight",
	"kolla/centos-source-neutron-sriov-agent",
	"kolla/centos-source-neutron-vpnaas-agent",
	"kolla/centos-source-nova-api",
	"kolla/centos-source-nova-compute",
	"kolla/centos-source-nova-compute-ironic",
	"kolla/centos-source-nova-conductor",
	"kolla/centos-source-nova-consoleauth",
	"kolla/centos-source-nova-libvirt",
	"kolla/centos-source-nova-novncproxy",
	"kolla/centos-source-nova-placement-api",
	"kolla/centos-source-nova-scheduler"]
}

修改部署配置文件
- 修改当前目录下的multinode文件：mutinode
- 修改/etc/kolla/global.yml文件：global.yml

部署：

生产随机密码文件：

kolla-genpwd

修改horizon登录界面admin密码：

[root@kolla ~]# vim /etc/kolla/passwords.yml
keepalived_password: mFbTVxF6XyrrT8NqaN5UpFB098GEXuZ9oQyfQI14
keystone_admin_password: admin  # 更改此处
keystone_database_password: C4EzIx0zhoFjsG9dA9TBRaZfbFIdT3f9sCe7jGyg

引导配置各节点依赖软件:

kolla-ansible -i ./multinode bootstrap-servers
PLAY RECAP *************************************************************************************************************************************************************
compute01                  : ok=38   changed=7    unreachable=0    failed=0   
compute02                  : ok=38   changed=7    unreachable=0    failed=0   
controller01               : ok=38   changed=7    unreachable=0    failed=0   
controller02               : ok=39   changed=17   unreachable=0    failed=0   
controller03               : ok=38   changed=7    unreachable=0    failed=0   
localhost                  : ok=1    changed=0    unreachable=0    failed=0

进行预部署检查：

kolla-ansible -i ./multinode prechecks
PLAY RECAP ************************************************************************************************************************************************************
compute01                  : ok=26   changed=1    unreachable=0    failed=0   
compute02                  : ok=26   changed=1    unreachable=0    failed=0   
controller01               : ok=91   changed=1    unreachable=0    failed=0   
controller02               : ok=87   changed=1    unreachable=0    failed=0   
controller03               : ok=87   changed=1    unreachable=0    failed=0   
localhost                  : ok=6    changed=1    unreachable=0    failed=0

Cinder出现错误

TASK [cinder : Checking LVM volume group exists for Cinder] ***********************************************************************************************************
skipping: [controller01]
skipping: [controller02]
skipping: [controller03]
[DEPRECATION WARNING]: Using tests as filters is deprecated. Instead of using `result|failed` use `result is failed`. This feature will be removed in version 2.9. 
Deprecation warnings can be disabled by setting deprecation_warnings=False in ansible.cfg.
fatal: [compute01]: FAILED! => {"changed": false, "cmd": ["vgs", "cinder-volumes"], "delta": "0:00:00.009794", "end": "2018-10-13 18:33:13.868282", "failed_when_result": true, "msg": "non-zero return code", "rc": 5, "start": "2018-10-13 18:33:13.858488", "stderr": "  Volume group \"cinder-volumes\" not found\n  Cannot process volume group cinder-volumes", "stderr_lines": ["  Volume group \"cinder-volumes\" not found", "  Cannot process volume group cinder-volumes"], "stdout": "", "stdout_lines": []}
[DEPRECATION WARNING]: Using tests as filters is deprecated. Instead of using `result|failed` use `result is failed`. This feature will be removed in version 2.9. 
Deprecation warnings can be disabled by setting deprecation_warnings=False in ansible.cfg.
fatal: [compute02]: FAILED! => {"changed": false, "cmd": ["vgs", "cinder-volumes"], "delta": "0:00:00.010114", "end": "2018-10-13 18:33:13.860281", "failed_when_result": true, "msg": "non-zero return code", "rc": 5, "start": "2018-10-13 18:33:13.850167", "stderr": "  Volume group \"cinder-volumes\" not found\n  Cannot process volume group cinder-volumes", "stderr_lines": ["  Volume group \"cinder-volumes\" not found", "  Cannot process volume group cinder-volumes"], "stdout": "", "stdout_lines": []}

* 解决方案：

[root@compute02 .ssh]# vgdisplay
  --- Volume group ---
  VG Name               centos
  System ID             
  Format                lvm2
  Metadata Areas        1
  Metadata Sequence No  4
  VG Access             read/write
  VG Status             resizable
  MAX LV                0
  Cur LV                3
  Open LV               3
  Max PV                0
  Cur PV                1
  Act PV                1
  VG Size               <1.82 TiB
  PE Size               4.00 MiB
  Total PE              476806
  Alloc PE / Size       476806 / <1.82 TiB
  Free  PE / Size       0 / 0   
  VG UUID               FEgDXH-SBlh-x29N-qU0f-Wajd-2sJ6-rbUre5
   
[root@compute02 .ssh]# dd if=/dev/zero of=./disk.img count=200 bs=512MB
200+0 records in
200+0 records out
102400000000 bytes (102 GB) copied, 509.072 s, 201 MB/s
[root@compute02 .ssh]# losetup -f
/dev/loop0
[root@compute02 .ssh]# losetup /dev/loop0 disk.img
[root@compute02 .ssh]# pvcreate /dev/loop0
  Physical volume "/dev/loop0" successfully created.
[root@compute02 .ssh]# vgcreate cinder-volumes /dev/loop0
  Volume group "cinder-volumes" successfully created

进行实际部署：

kolla-ansible -i ./multinode deploy

初始化OpenStack

删除ipadress的py包并重新安装（版本过低下一步客户端安装会出错，原先安装其他包的时候作为依赖包安装的ipaddress无法通过pip删除并升级，只能手动删除再安装最新版本）：

[root@kolla ~]# cd /usr/lib/python2.7/site-packages/
[root@kolla site-packages]# rm -rf ipaddress*
[root@kolla site-packages]# pip install ipaddress

安装OpenStack CLI客户端:

[root@kolla site-packages]# pip install python-openstackclient python-glanceclient python-neutronclient

设置环境变量：

[root@kolla site-packages]# . /etc/kolla/admin-openrc.sh

编辑初始化脚本中的网络配置：

[root@kolla ~]# vim /usr/share/kolla-ansible/init-runonce
EXT_NET_CIDR='10.132.226.0/24'
EXT_NET_RANGE='start=10.132.226.130,end=10.132.226.169'
EXT_NET_GATEWAY='10.132.226.254'

执行初始化脚本：

[root@kolla ~]# . /usr/share/kolla-ansible/init-runonce
Checking for locally available cirros image.
None found, downloading cirros image.
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                             Dload  Upload   Total   Spent    Left  Speed
100 12.1M  100 12.1M    0     0  2040k      0  0:00:06  0:00:06 --:--:-- 2716k
Creating glance image.
······
Done.

To deploy a demo instance, run:

openstack server create \
    --image cirros \
    --flavor m1.tiny \
    --key-name mykey \
    --nic net-id=89a1f674-e89f-4e6d-b96d-2875446adc1e \
    demo1

Dolphinsz

关注

0
点赞
踩
11

收藏

觉得还不错? 一键收藏
2
评论
kolla-ansible 部署OpenStack queens版本笔记

kolla-ansible 部署OpenStack queens版本笔记一. 实验环境：6台主机安装CentOS7 minimal系统32G内存，1T+500G双硬盘（其中一个为后期ceph部署做准备），三张千兆网卡（一张用于后期ceph（enp3s0f0），另两张网卡分别作为控制网络（enp0s31f6）以及neutron桥接（enp3s0f1）网络）网络规划：hostIP...
复制链接

扫一扫

专栏目录