前言
这部分我们定制一个“经典网络”，即提供对内和对外的网络模型。对外由provider网络提供，对内由internal网络提供。
一、准备工作
在VMware的control和compute主机上添加仅主机模型的网卡。
添加完成后启动control和compute主机，然后分别在两个主机中配置新网卡eth1。
cd /etc/sysconfig/network-scripts
cp ifcfg-eth0 ifcfg-eth1
vim ifcfg-eth1
NAME=eth1
DEVICE=eth1
IPADDR=192.168.162.14 # 控制节点 ip 地址设置为 192.168.162.15
NETMASK=255.255.255.0注意：因为是仅主机模式的网卡，这里不需要设置Gateway和DNS。
设置完成后启动eth1网络。
ifup eth1二、定制网络
1. 控制节点配置
- 添加内网标识：
vim /etc/neutron/plugins/ml2/ml2_conf.ini
[ml2_type_flat]
flat_networks = provider,internal- 设置内网使用 eth1 进行通信：
vim /etc/neutron/plugins/ml2/linuxbridge_agent.ini
[linux_bridge]
physical_interface_mappings = provider:eth0,internal:eth1设置完成后启动网络相关服务。
systemctl restart neutron-server neutron-linuxbridge-agent
systemctl status neutron-server neutron-linuxbridge-agent2. 计算节点配置
计算节点只需要修改linuxbridge_agent.ini文件即可。
vim /etc/neutron/plugins/ml2/linuxbridge_agent.ini
[linux_bridge]
physical_interface_mappings = provider:eth0,internal:eth1设置完成后重启服务。
systemctl restart neutron-linuxbridge-agent
systemctl status neutron-linuxbridge-agent3. 创建internal网络
- 检查网络现状和服务是否正常
- 创建internal网络
- 创建internal子网
# 在control节点上查看网络现状和服务是否正常
source openstack-admin.sh
openstack network agent list
openstack compute service list
openstack network list
# 创建internal网络，网络类型选择flat
openstack network create --share --internal --provider-physical-network internal --provider-network-type flat internal
# 创建子网
openstack subnet create --network internal --allocation-pool start=192.168.162.100,end=192.168.162.120 --dns-nameserver 223.5.5.5 --gateway 192.168.162.2 --subnet-range 192.168.162.0/24 internal
# 检查
openstack network list4. 开启双网卡功能
1）首先在任意一台虚拟机中绑定上面的internal网络，具体操作可以在dashboard完成；
2）ssh连接该虚拟机；
3）开启支持双网卡的功能；
sudo su -
vi /etc/network/interfaces
# 追加以下两条配置，表示支持eth1启动
auto eth1
iface eth1 inet dhcp配置完成后，重新启动eth1网络即可。
ifup eth1
ip addr li eth1
三、节点扩容
下面演示如何增加一个新的计算节点。
1. 准备工作
- 检查网络以及相关服务是否正常
- 修改源
- 安装nova相关软件
- 安装网络软件
# 检查网络和服务是否正常
source openstack-admin.sh
openstack compute service list
openstack network agent list
# 修改源
scp [email protected]:/etc/yum.repos.d/CentOS-Base-163.repo /etc/yum.repos.d
# 安装nova
yum install -y openstack-nova-compute --nogpgchec
# 安装网络软件
yum install -y openstack-neutron-linuxbridge --nogpgchec
yum install -y conntrack-tools --nogpgchec2. 配置nova
- 配置网络
- 开启nova-api功能
- 配置rabbitmq地址
- 配置keystone认证
- 配置kvm
- 配置资源跟踪
- 配置vnc
vim /etc/nova/nova.conf
[DEFAULT]
use_neutron = True
firewall_driver = nova.virt.firewall.NoopFirewallDriver
enabled_apis = osapi_compute,metadata
transport_url = rabbit://openstack:[email protected]:5672
[api]
auth_strategy = keystone
[glance]
api_servers = http://192.168.88.14:9292
[keystone_authtoken]
www_authenticate_uri = http://192.168.88.14:5000
auth_url = http://192.168.88.14:35357
memcached_servers = 192.168.88.14:11211
auth_type = password
project_domain_name = default
user_domain_name = default
project_name = service
username = nova
password = nova
[libvirt]
virt_type = qemu
vif_plugging_is_fatal=false
vif_plugging_timeout=10
[oslo_concurrency]
lock_path = /var/lib/nova/tmp
[placement]
os_region_name = RegionOne
project_domain_name = Default
project_name = service
auth_type = password
user_domain_name = Default
auth_url = http://192.168.88.14:35357/v3
username = placement
password = placement
[vnc]
enabled = true
vncserver_listen = 0.0.0.0
vncserver_proxyclient_address = 192.168.88.15
novncproxy_base_url = http://192.168.88.14:6080/vnc_auto.html配置完成后重启nova相关服务。
systemctl start libvirtd openstack-nova-compute
systemctl status libvirtd openstack-nova-compute
systemctl enable libvirtd openstack-nova-compute最后，验证服务是否正常。
# 切换admin用户
source openstack-admin.sh
# 检查nova-compute服务是否正常
openstack compute service list
# 数据库增加计算节点记录
su -s /bin/sh -c "nova-manage cell_v2 discover_hosts --verbose" nova
# 验证所有API是否正常
nova-status upgrade check3. 配置neutron
- 将网卡配置文件中增加ipv6的功能
- 配置keystone认证
- 配置rabbitmq地址
- 配置锁文件
# 将网卡配置文件中增加ipv6的功能：
vim /etc/sysctl.conf
net.ipv6.conf.all.disable_ipv6 = 0
net.ipv6.conf.default.disable_ipv6 = 0
net.ipv6.conf.lo.disable_ipv6 = 0
# 永久生效
echo 'sysctl -p' >> /etc/profile
vim /etc/neutron/neutron.conf
[DEFAULT]
auth_strategy = keystone
transport_url = rabbit://openstack:[email protected]:5672
[keystone_authtoken]
auth_uri = http://192.168.88.14:5000
auth_url = http://192.168.88.14:35357
memcached_servers = 192.168.88.14:11211
auth_type = password
project_domain_name = default
user_domain_name = default
project_name = service
username = neutron
password = neutron
[oslo_concurrency]
lock_path = /var/lib/neutron/tmp4. 配置网桥
- 配置使用的物理网卡
- 关闭vxlan
- 配置安全组
vim /etc/neutron/plugins/ml2/linuxbridge_agent.ini
[linux_bridge]
physical_interface_mappings = provider:eth0
[vxlan]
enable_vxlan = false
[securitygroup]
enable_security_group = true
firewall_driver = neutron.agent.linux.iptables_firewall.IptablesFirewallDriver 5. 修改nova配置
- 添加网络配置
vim /etc/nova/nova.conf
[neutron]
url = http://192.168.88.14:9696
auth_url = http://192.168.88.14:35357
auth_type = password
project_domain_name = default
user_domain_name = default
region_name = RegionOne
project_name = service
username = neutron
password = neutron配置完成后，重启相关服务。
# 重启openstack-nova-compute
systemctl restart openstack-nova-compute
systemctl status openstack-nova-compute
# 启动neutron-linuxbridge-agent
systemctl start neutron-linuxbridge-agent
systemctl status neutron-linuxbridge-agent
systemctl enable neutron-linuxbridge-agent最后，验证服务是否正常。
# 查看所有的网络
openstack network agent list
# 查看所有的计算节点服务
openstack compute service list四、删除节点
1. 删除节点
- 查看要删除计算节点的ID
- 删除计算节点
- 停止nova相关服务
# 查看所有计算节点
openstack compute service list
+----+----------------+----------+----------+----------+-------+----------------------------+
| ID | Binary | Host | Zone | Status | State | Updated At |
+----+----------------+----------+----------+----------+-------+----------------------------+
| 5 | nova-console | control | internal | enabled | up | 2021-04-02T10:05:44.000000 |
| 6 | nova-scheduler | control | internal | enabled | up | 2021-04-02T10:05:43.000000 |
| 7 | nova-conductor | control | internal | enabled | up | 2021-04-02T10:05:44.000000 |
| 8 | nova-compute | compute | nova | enabled | down | 2021-04-02T07:38:49.000000 |
| 9 | nova-compute | image | nova | disabled | down | 2021-04-02T09:58:55.000000 |
| 10 | nova-compute | compute2 | nova | enabled | up | 2021-04-02T10:05:39.000000 |
+----+----------------+----------+----------+----------+-------+----------------------------+
# 选择要删除的计算节点
openstack compute service delete 10
# 在计算节点上停止相关服务
systemctl stop libvirtd openstack-nova-compute2. 清除网络信息
# 查询网络节点ID
openstack network agent list
# 删除网络节点
openstack network agent delete <agent_id>
# 停止服务
systemctl stop neutron-linuxbridge-agent3. 清除数据库信息
# 登录数据库
mysql -unova -pnova
use nova
# 删除计算记录
select host from nova.services;
delete from nova.services where host="compute2";
# 删除网络记录
select hypervisor_hostname from compute_nodes;
delete from compute_nodes where hypervisor_hostname="compute2";
exit;最后检查一下：
openstack compute service list
openstack network agent list总结
这部分介绍了如何实现经典网络的双网卡功能。虽然现在经典网络已经过时，但是对于openstack新手而言，是一个不错的学习材料。除此以外，我们还了解如何对节点进行扩容和缩容。到目前为止，整个openstack的最小化部署已经基本完成。