天楚锐齿

人工智能 云计算 大数据 物联网 IT 通信 嵌入式

OpenStack 安装和使用（三）

使用Compute
安装：
# yum install –enablerepo=updates-testing libnova
# yum install –enablerepo=updates-testing libnova-devel
# yum install –enablerepo=updates-testing openstack-nova-doc
# yum install –enablerepo=updates-testing openstack-nova-novncproxy
# yum install –enablerepo=updates-testing python-nova-adminclient
# yum install –enablerepo=updates-testing python-novaclient-doc
# yum install –enablerepo=updates-testing novacom
修改配置：
# nova-manage db sync
# openstack-config –set /etc/nova/api-paste.ini filter:authtoken admin_tenant_name service
# openstack-config –set /etc/nova/api-paste.ini filter:authtoken admin_user nova
# openstack-config –set /etc/nova/api-paste.ini filter:authtoken admin_password nova
# openstack-config –set /etc/nova/nova.conf DEFAULT auth_strategy keystone
# openstack-config –set /etc/nova/nova.conf DEFAULT connection_type libvirt
# openstack-config –set /etc/nova/nova.conf DEFAULT libvirt_type kvm
# openstack-config –set /etc/nova/nova.conf DEFAULT multi_host True
# openstack-config –set /etc/nova/nova.conf DEFAULT enabled_apis ec2,osapi_compute,osapi_volume,metadata
# openstack-config –set /etc/nova/nova.conf DEFAULT volume_group vg_nova_volume
# openstack-config –set /etc/nova/nova.conf DEFAULT force_dhcp_release False
# openstack-config –set /etc/nova/nova.conf DEFAULT libvirt_inject_partition -1
# openstack-config –set /etc/nova/nova.conf DEFAULT qpid_hostname 10.24.1.47                                       为qpid节点地址

# openstack-config –set /etc/nova/nova.conf DEFAULT sql_connection mysql://nova:nova@10.24.1.47/nova   为mysql所在的地址
# openstack-config –set /etc/nova/nova.conf DEFAULT glance_api_servers 10.24.1.47:9292                          为glance api节点地址
# openstack-config –set /etc/nova/nova.conf DEFAULT iscsi_ip_prefix 10.24.1.47                                         为cloud controller节点地址

检查物理机是否支持虚拟化：
# egrep ‘(vmx|svm)’ –color=always /proc/cpuinfo
可以看到红字的svm或vmx标记，可能要从bios里面打开。
# lsmod | grep kvm
kvm_amd                40318  0
kvm                   305081  1 kvm_amd
# modprobe kvm
# modprobe kvm_amd
# modprobe nbd
启动服务：
# for svc in api objectstore compute network volume scheduler cert console consoleauth direct-api novncproxy xvpvncproxy ; do systemctl enable openstack-nova-$svc.service ;  done
# killall dnsmasq; for svc in api objectstore compute network volume scheduler cert console consoleauth direct-api novncproxy xvpvncproxy ; do systemctl restart openstack-nova-$svc.service ; done
# for svc in api objectstore compute network volume scheduler cert console consoleauth direct-api novncproxy xvpvncproxy ; do systemctl stop openstack-nova-$svc.service ; done
# for svc in api objectstore compute network volume scheduler cert console consoleauth direct-api novncproxy xvpvncproxy ; do systemctl status openstack-nova-$svc.service ; done

# systemctl -a –full –no-page list-units |grep nova

# systemctl -a –full –no-page list-unit-files |grep nova
novacomd.service                            disabled
openstack-nova-api.service                  enabled
openstack-nova-cert.service                 enabled
openstack-nova-compute.service              enabled
openstack-nova-console.service              enabled
openstack-nova-consoleauth.service          enabled
openstack-nova-direct-api.service           enabled
openstack-nova-metadata-api.service         disabled
openstack-nova-network.service              enabled
openstack-nova-novncproxy.service           enabled
openstack-nova-objectstore.service          enabled
openstack-nova-scheduler.service            enabled
openstack-nova-volume.service               enabled

openstack-nova-xvpvncproxy.service          enabled
这里面的metadata-api可能不能启动，是因为启动了api导致的，api含有metadata-api，metadata-api可以用于纯粹的compute节点(不含cloud controller的功能)。
管理节点或代理节点需要启动上面所有的进程(除了metadata-pai)，纯粹的计算节点只需要：nova-compte, nova-network, nova-api或metadata-api。
建立VM需要用到的网络：
# nova-manage network create hanborq –multi_host=T –fixed_range_v4=10.25.1.0/24 –gateway=10.25.1.1 –dns1=8.8.8.8 –bridge=hanborqbr0 –bridge_interface=em2 –num_networks=1 –network_size=256
# nova-manage network list
id      IPv4                    IPv6            start address   DNS1            DNS2            VlanID          project         uuid
1       10.25.1.0/24            None            10.25.1.2       8.8.8.8         None            None            None            37f4503b-58b0-4773-8001-4e95942b185c
注意：
1. 最好不要使用nova-manage netowrk delete命令，而是使用nova-manage network modify来修改。
2. 最好不要使用已有网段，FLAT DHCP网络方式时要确保不跟别的ip地址冲突，另外DHCP方式最好不要使用管理用的eth0(em1)来作网桥，这里用eth1(em2)。
校验NOVA：

# nova-manage version list

2012.1 (2012.1-LOCALBRANCH:LOCALREVISION)
# nova flavor-list
+—-+———–+———–+——+———–+——+——-+————-+
| ID |    Name   | Memory_MB | Disk | Ephemeral | Swap | VCPUs | RXTX_Factor |
+—-+———–+———–+——+———–+——+——-+————-+
| 1  | m1.tiny   | 512       | 0    | 0         |      | 1     | 1.0         |
| 2  | m1.small  | 2048      | 10   | 20        |      | 1     | 1.0         |
| 3  | m1.medium | 4096      | 10   | 40        |      | 2     | 1.0         |
| 4  | m1.large  | 8192      | 10   | 80        |      | 4     | 1.0         |
| 5  | m1.xlarge | 16384     | 10   | 160       |      | 8     | 1.0         |
+—-+———–+———–+——+———–+——+——-+————-+
# nova-manage service list
Binary           Host                                 Zone             Status     State Updated_At
nova-console     cc                                   nova             enabled    🙂   2012-07-07 03:28:43
nova-consoleauth cc                                   nova             enabled    🙂   2012-07-07 03:28:42
nova-compute     cc                                   nova             enabled    🙂   2012-07-07 03:28:39
nova-scheduler   cc                                   nova             enabled    🙂   2012-07-07 03:28:43
nova-cert        cc                                   nova             enabled    🙂   2012-07-07 03:28:43
nova-network     cc                                   nova             enabled    🙂   2012-07-07 03:28:36
nova-volume      cc                                   nova             enabled    🙂   2012-07-07 03:28:41
# nova image-list
+————————————–+——————-+——–+——–+
|                  ID                  |        Name       | Status | Server |
+————————————–+——————-+——–+——–+
| 2eb1ed9e-b1f2-450a-8ce2-810580d09610 | tty-linux-kernel  | ACTIVE |        |
| e36a1d2d-7ae1-4496-b25c-3ee4be0a14d8 | tty-linux-ramdisk | ACTIVE |        |
| e951ba9a-a5f1-40fb-916a-56fc2aeee429 | tty-linux         | ACTIVE |        |
+————————————–+——————-+——–+——–+
安装其他compute节点：
cloud controller节点需要开放amqp、MySQL、Nova API、iSCS等I端口，上面lokkit命令已经把他们开放。
跟cloud controller一样启动ntp服务。
# yum install –enablerepo=updates-testing ntp
# ntpdate 0.fedora.pool.ntp.org
# service ntpd start
# chkconfig ntpd on
安装openstack-nova组件。
# yum install –enablerepo=updates-testing openstack-nova
# service libvirtd start
# chkconfig libvirtd on
# setenforce 0
# getenforce
Permissive
修改配置文件api-paste.ini和nova.conf：
# openstack-config –set /etc/nova/nova.conf DEFAULT auth_strategy keystone
# openstack-config –set /etc/nova/nova.conf DEFAULT libvirt_inject_partition -1
# openstack-config –set /etc/nova/nova.conf DEFAULT connection_type libvirt
# openstack-config –set /etc/nova/nova.conf DEFAULT libvirt_type kvm
# openstack-config –set /etc/nova/nova.conf DEFAULT multi_host True
# openstack-config –set /etc/nova/nova.conf DEFAULT enabled_apis metadata               注意跟cc节点是不一样的

# openstack-config –set /etc/nova/nova.conf DEFAULT force_dhcp_release False

# openstack-config –set /etc/nova/nova.conf DEFAULT qpid_hostname 10.24.1.47                                       为qpid节点地址
# openstack-config –set /etc/nova/nova.conf DEFAULT sql_connection mysql://nova:nova@10.24.1.47/nova   为mysql所在的地址
# openstack-config –set /etc/nova/nova.conf DEFAULT glance_api_servers 10.24.1.47:9292                          为glance api节点地址
# openstack-config –set /etc/nova/nova.conf DEFAULT iscsi_ip_prefix 10.24.1.47                                         为cloud controller节点地址
启动compute和network进程：
# killall dnsmasq; for svc in api compute network ; do systemctl restart openstack-nova-$svc.service ; done
# for svc in api compute network ; do systemctl status openstack-nova-$svc.service ; done
# for svc in api compute network ; do systemctl enable openstack-nova-$svc.service ; done
# systemctl -a –full –no-page list-unit-files |grep nova
# systemctl -a –full –no-page list-units |grep nova
openstack-nova-compute.service                                                                                             loaded active   running       OpenStack Nova Compute Server
openstack-nova-metadata-api.service                                                                                        loaded active   running       OpenStack Nova Metadata API Server
openstack-nova-network.service                                                                                             loaded active   running       OpenStack Nova Network Server
设置环境变量：
# vi .bash_profile
…
export ADMIN_TOKEN=c8701a128708a76161e3
export OS_USERNAME=admin
export OS_PASSWORD=123456
export OS_TENANT_NAME=hanborq
export OS_AUTH_URL=http://10.24.1.47:5000/v2.0/
看看客户端是否好使：
# nova-manage service list
# nova image-list
# nova list
注册和运行虚拟机
用f16jeos来进行测试：
增加image：
# glance add name=f16-jeos is_public=true disk_format=qcow2 container_format=bare < f16-x86_64-openstack-sda.qcow2
磁盘格式disk_format可以用# file f16-x86_64-openstack-sda.qcow2 看出来(有iso, qcow2, raw等)，container_format目前没有使用，可以是bare, ovf, aki, ari, ami，所以所有image都可以用bare。
# glance index
ID                                   Name                           Disk Format          Container Format     Size
———————————— —————————— ——————– ——————– ————–
3e00b8a0-5159-4f60-b757-f00273d65c37 f16-jeos                       qcow2                bare                      213581824
e951ba9a-a5f1-40fb-916a-56fc2aeee429 tty-linux                      ami                  ami                        25165824
e36a1d2d-7ae1-4496-b25c-3ee4be0a14d8 tty-linux-ramdisk              ari                  ari                           96629
2eb1ed9e-b1f2-450a-8ce2-810580d09610 tty-linux-kernel               aki                  aki                         4404752
# nova image-list
+————————————–+———————+——–+——–+
|                  ID                  |         Name        | Status | Server |
+————————————–+———————+——–+——–+
| 3e00b8a0-5159-4f60-b757-f00273d65c37 | f16-jeos          | ACTIVE |        |
| 2eb1ed9e-b1f2-450a-8ce2-810580d09610 | tty-linux-kernel    | ACTIVE |        |
| e36a1d2d-7ae1-4496-b25c-3ee4be0a14d8 | tty-linux-ramdisk   | ACTIVE |        |
| e951ba9a-a5f1-40fb-916a-56fc2aeee429 | tty-linux           | ACTIVE |        |
+————————————–+———————+——–+——–+
设置安全组，nova使用安全组来控制哪些地址、端口、协议可以被访问，默认具有default组：
# nova secgroup-list
+———+————-+
|   Name  | Description |
+———+————-+
| default | default     |
+———+————-+
往default组里面添加规则，允许所有地址来访问VM的22 TCP端口(ssh)：
# nova secgroup-add-rule default tcp 22 22 0.0.0.0/0
+————-+———–+———+———–+————–+
| IP Protocol | From Port | To Port |  IP Range | Source Group |
+————-+———–+———+———–+————–+
| tcp         | 22        | 22      | 0.0.0.0/0 |              |
+————-+———–+———+———–+————–+
允许所有地址来访问VM的ICMP协议（ping）：
# nova secgroup-add-rule default icmp -1 -1 0.0.0.0/0
+————-+———–+———+———–+————–+
| IP Protocol | From Port | To Port |  IP Range | Source Group |
+————-+———–+———+———–+————–+
| icmp        | -1        | -1      | 0.0.0.0/0 |              |
+————-+———–+———+———–+————–+
显示default组下面的规则：
# nova secgroup-list-rules default
+————-+———–+———+———–+————–+
| IP Protocol | From Port | To Port |  IP Range | Source Group |
+————-+———–+———+———–+————–+
| icmp        | -1        | -1      | 0.0.0.0/0 |              |
| tcp         | 22        | 22      | 0.0.0.0/0 |              |
+————-+———–+———+———–+————–+
生成ssh鉴权需要用的key值：
# nova keypair-add f16jeos > f16jeos.priv
# chmod 600 f16jeos.priv
# vi f16jeos.priv
# nova keypair-list+———+————————————————-+

|   Name  |                   Fingerprint                   |
+———+————————————————-+
| f16jeos | b1:e0:7c:1d:f8:54:63:fd:dd:4b:b2:4b:ff:6b:52:17 |
+———+————————————————-+

查看具有哪些默认的Flavor（既虚拟机的硬件配置）：
# nova flavor-list
+—-+———–+———–+——+———–+——+——-+————-+
| ID |    Name   | Memory_MB | Disk | Ephemeral | Swap | VCPUs | RXTX_Factor |
+—-+———–+———–+——+———–+——+——-+————-+
| 1  | m1.tiny   | 512       | 0    | 0         |      | 1     | 1.0         |
| 2  | m1.small  | 2048      | 10   | 20        |      | 1     | 1.0         |
| 3  | m1.medium | 4096      | 10   | 40        |      | 2     | 1.0         |
| 4  | m1.large  | 8192      | 10   | 80        |      | 4     | 1.0         |
| 5  | m1.xlarge | 16384     | 10   | 160       |      | 8     | 1.0         |
+—-+———–+———–+——+———–+——+——-+————-+
启动虚拟机：
其中flavor为2标识m1.small类型，image为刚才上传的cirros的image id，key_name为刚才生成的ssh key的名字，安全组为default，虚拟机的名称为cirros_vm，至于虚拟机跑在哪个物理主机上，由scheduler(有好多配置可选)决定，也可以由”–hint force_hosts=cc“直接指定：
如果要指定多个网卡，则需要增加“-nic net-id=XXXX -nic net-id=YYYY“，这里XXXX和YYYY为上面建立的network id。
# glance index
# nova boot –flavor 2 –image 3e00b8a0-5159-4f60-b757-f00273d65c37 –key_name f16jeos –security_group default f16jeos_vm
+————————————-+————————————–+
|               Property              |                Value                 |
+————————————-+————————————–+
| OS-DCF:diskConfig                   | MANUAL                               |
| OS-EXT-SRV-ATTR:host                | None                                 |
| OS-EXT-SRV-ATTR:hypervisor_hostname | None                                 |
| OS-EXT-SRV-ATTR:instance_name       | instance-00000001                    |
| OS-EXT-STS:power_state              | 0                                    |
| OS-EXT-STS:task_state               | scheduling                           |
| OS-EXT-STS:vm_state                 | building                             |
| accessIPv4                          |                                      |
| accessIPv6                          |                                      |
| adminPass                           | v8faNaee8XNv                         |
| config_drive                        |                                      |
| created                             | 2012-07-16T10:19:28Z                 |
| flavor                              | m1.small                             |
| hostId                              |                                      |
| id                                  | cb3f608d-317b-4088-9a0a-53fdfc4edf80 |
| image                               | f16-jeos                             |
| key_name                            | f16jeos                              |
| metadata                            | {}                                   |
| name                                | f16jeos_vm                           |
| progress                            | 0                                    |
| status                              | BUILD                                |
| tenant_id                           | 94d38db32a7d4107beeed36d9e98bf06     |
| updated                             | 2012-07-16T10:19:28Z                 |
| user_id                             | cd5e2e764fc7420dad016f9af7be7123     |
+————————————-+————————————–+
查看虚拟机：
# nova list
+————————————–+————+——–+——————-+

|                  ID                  |    Name    | Status |      Networks     |
+————————————–+————+——–+——————-+
| cb3f608d-317b-4088-9a0a-53fdfc4edf80 | f16jeos_vm | ACTIVE | hanborq=10.25.1.2 |
+————————————–+————+——–+——————-+

# nova show f16jeos_vm
+————————————-+———————————————————-+
|               Property              |                          Value                           |
+————————————-+———————————————————-+
| OS-DCF:diskConfig                   | MANUAL                                                   |
| OS-EXT-SRV-ATTR:host                | compute1                                                 |
| OS-EXT-SRV-ATTR:hypervisor_hostname | None                                                     |
| OS-EXT-SRV-ATTR:instance_name       | instance-00000011                                        |
| OS-EXT-STS:power_state              | 1                                                        |
| OS-EXT-STS:task_state               | None                                                     |
| OS-EXT-STS:vm_state                 | active                                                   |
| accessIPv4                          |                                                          |
| accessIPv6                          |                                                          |
| config_drive                        |                                                          |
| created                             | 2012-07-18T08:05:22Z                                     |
| flavor                              | m1.small                                                 |
| hanborq network                     | 10.25.1.4                                                |
| hostId                              | ba210c21c18c8c89c1848c6fb140ac5936c5d539ea17737a2921b7e2 |
| id                                  | 81decfae-c4dd-41c7-83dd-7bc8f03fd897                     |
| image                               | f16-jeos                                                 |
| key_name                            | f16jeos                                                  |
| metadata                            | {}                                                       |
| name                                | f16jeos_vm_1                                             |
| progress                            | 0                                                        |
| status                              | ACTIVE                                                   |
| tenant_id                           | 94d38db32a7d4107beeed36d9e98bf06                         |
| updated                             | 2012-07-18T08:05:37Z                                     |
| user_id                             | cd5e2e764fc7420dad016f9af7be7123                         |
+————————————-+———————————————————-+
查看虚拟机console内容：
# virsh list
Id    Name                           State
—————————————————-
1     instance-00000001              running
# nova console-log f16jeos_vm
# nova delete f16jeos_vm
# ping 10.25.1.2
# ssh -i f16jeos.priv ec2-user@10.25.1.2
[ec2-user@f16jeos-vm ~]$ exit
直接登录到虚拟机，虚拟机的网络地址为10.25.1.2，接口名字为eth0，从下面物理主机的ifconfig命令可以看到该接口对应的物理主机接口为vnet0。
# ifconfig
可以看到多出来了hanborqbr0和vnet0两个接口，其中hanborqbr0为网桥接口(顺便做虚拟机的DHCP server)，vnet0为虚拟机内部的网口，每加一个虚拟机则加一个vnetX。
em1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
inet 10.24.1.47  netmask 255.255.255.0  broadcast 10.24.1.255
inet6 fe80::222:19ff:fe5d:4847  prefixlen 64  scopeid 0x20<link>
ether 00:22:19:5d:48:47  txqueuelen 1000  (Ethernet)
RX packets 329574  bytes 417245559 (397.9 MiB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 194639  bytes 17563292 (16.7 MiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
device interrupt 37  memory 0xec000000-ec012800

em2: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
inet6 fe80::222:19ff:fe5d:4849  prefixlen 64  scopeid 0x20<link>
ether 00:22:19:5d:48:49  txqueuelen 1000  (Ethernet)
RX packets 30338  bytes 2746833 (2.6 MiB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 37  bytes 5733 (5.5 KiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
device interrupt 37  memory 0xea000000-ea012800

hanborqbr0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
inet 10.25.1.3  netmask 255.255.255.0  broadcast 10.25.1.255
inet6 fe80::4a9:54ff:fef2:82ca  prefixlen 64  scopeid 0x20<link>
ether 00:22:19:5d:48:49  txqueuelen 0  (Ethernet)
RX packets 466  bytes 38146 (37.2 KiB)
RX errors 0  dropped 3  overruns 0  frame 0
TX packets 169  bytes 19322 (18.8 KiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 16436
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10<host>
loop  txqueuelen 0  (Local Loopback)
RX packets 37409  bytes 459580998 (438.2 MiB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 37409  bytes 459580998 (438.2 MiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

virbr0: flags=4099<UP,BROADCAST,MULTICAST>  mtu 1500
inet 192.168.122.1  netmask 255.255.255.0  broadcast 192.168.122.255
ether 1a:bc:b0:89:bd:7f  txqueuelen 0  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

vnet0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
inet6 fe80::fc16:3eff:fe77:8918  prefixlen 64  scopeid 0x20<link>
ether fe:16:3e:77:89:18  txqueuelen 500  (Ethernet)
RX packets 148  bytes 14938 (14.5 KiB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 359  bytes 35029 (34.2 KiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

1,021次阅读

发表回复

您的邮箱地址不会被公开。 必填项已用 * 标注