本文共 10285 字，大约阅读时间需要 34 分钟。

先说出我找见错误是在哪里，虽然是很小的失误，但还是得注意

1，一次不小心把控制节点的主机名给修改了，然后就导致keystone工作不正常，消息机制出现错误。把控制节点的主机名改回来到原来的主机名就可以了。 2，创建实例出现错误： 查看nova-compute.log 看到的是：

Starting instan

ce… 2017-05-16 15:10:55.461 3677 WARNING nova.compute.resource_tracker [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Host field should not be set on the instance unt il resources have been claimed. 2017-05-16 15:10:55.461 3677 WARNING nova.compute.resource_tracker [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Node field should not be set on the instance unt il resources have been claimed. 2017-05-16 15:10:55.464 3677 AUDIT nova.compute.claims [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Attempting claim: memory 512 MB, disk 1 GB 2017-05-16 15:10:55.464 3677 AUDIT nova.compute.claims [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Total memory: 9891 MB, used: 8192.00 MB 2017-05-16 15:10:55.465 3677 AUDIT nova.compute.claims [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] memory limit not specified, defaulting to unlimited 2017-05-16 15:10:55.465 3677 AUDIT nova.compute.claims [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Total disk: 196 GB, used: 93.00 GB 2017-05-16 15:10:55.466 3677 AUDIT nova.compute.claims [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] disk limit not specified, defaulting to unlimited 2017-05-16 15:10:55.482 3677 AUDIT nova.compute.claims [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Claim successful 2017-05-16 15:10:55.818 3677 INFO nova.scheduler.client.report [-] Compute_service record updated for (‘sles11sp3x64-hsm1-vm3’, ‘sles11sp3x64-hsm1-vm3.istl.chd.edu.cn’) 2017-05-16 15:10:55.871 3677 INFO nova.network.neutronv2.api [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Unable to reset device ID for port None 2017-05-16 15:10:56.350 3677 INFO nova.scheduler.client.report [-] Compute_service record updated for (‘sles11sp3x64-hsm1-vm3’, ‘sles11sp3x64-hsm1-vm3.istl.chd.edu.cn’) 2017-05-16 15:10:56.657 3677 INFO nova.virt.libvirt.driver [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Creating image 2017-05-16 15:10:56.752 3677 INFO nova.virt.disk.vfs.api [-] Unable to import guestfsfalling back to VFSLocalFS 2017-05-16 15:10:56.754 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:10:56.754 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:10:56.943 3677 INFO nova.scheduler.client.report [-] Compute_service record updated for (‘sles11sp3x64-hsm1-vm3’, ‘sles11sp3x64-hsm1-vm3.istl.chd.edu.cn’) 2017-05-16 15:10:58.755 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:10:58.756 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:00.758 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:00.759 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:02.762 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:02.763 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:04.766 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:04.767 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:06.769 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:06.770 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:08.771 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:08.772 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:10.774 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:10.775 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:10.775 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:12.778 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:12.779 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:14.781 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:14.782 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:16.784 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:16.785 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:18.786 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:18.788 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:20.790 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:20.791 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:22.793 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:22.794 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:24.795 3677 ERROR nova.virt.disk.mount.nbd [-] nbd module not loaded 2017-05-16 15:11:24.796 3677 INFO nova.virt.disk.mount.api [-] Device allocation failed. Will retry in 2 seconds. 2017-05-16 15:11:26.797 3677 WARNING nova.virt.disk.mount.api [-] Device allocation failed after repeated retries. 2017-05-16 15:11:26.798 3677 INFO nova.virt.libvirt.driver [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Using config drive 2017-05-16 15:11:26.892 3677 INFO nova.virt.libvirt.driver [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Creating config drive at /var/lib/nova/instances/9cd465f 8-76ec-40c2-8222-04796818f552/disk.config 2017-05-16 15:11:26.987 3677 ERROR nova.compute.manager [-] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Instance failed to spawn 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Traceback (most recent call last): 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] File “/usr/lib64/python2.6/site-packages/nova/compute/manager .py”, line 2317, in _build_resources 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] yield resources 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] File “/usr/lib64/python2.6/site-packages/nova/compute/manager .py”, line 2187, in _build_and_run_instance 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] block_device_info=block_device_info) 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] File “/usr/lib64/python2.6/site-packages/nova/virt/libvirt/driver.py”, line 2682, in spawn 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] write_to_disk=True) 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] File “/usr/lib64/python2.6/site-packages/nova/virt/libvirt/dr iver.py”, line 4212, in _get_guest_xml 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] context) 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] File “/usr/lib64/python2.6/site-packages/nova/virt/libvirt/dr iver.py”, line 3990, in _get_guest_config 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] flavor, CONF.libvirt.virt_type) 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] File “/usr/lib64/python2.6/site-packages/nova/virt/libvirt/vif.py”, line 358, in get_config 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] _(“Unexpected vif_type=%s”) % vif_type) 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] NovaException: Unexpected vif_type=binding_failed 2017-05-16 15:11:26.987 3677 TRACE nova.compute.manager [instance: 9cd465f8-76ec-40c2-8222-04796818f552] 2017-05-16 15:11:26.989 3677 AUDIT nova.compute.manager [req-50e1bc3a-ebb8-4ec5-979d-c3db0f0e2fe0 None] [instance: 9cd465f8-76ec-40c2-8222-04796818f552] Terminating ins Tance

首先我们要做的是：

1、检查openstack的各个组件是否正常工作 查看nova组件：

命令：nova service-list

查看neutron组件是否正常工作：

命令：neutron agent-list 感觉是这个neutron-openvswitch-agent出问题了

neutron-openvswitch-agent只是Neutron的一个网络插件，那么什么是openvswitch，这里我们需要了解下：

OpenvSwitch，简称OVS是一个虚拟交换软件，主要用于虚拟机VM环境，作为一个虚拟交换机，支持Xen/XenServer, KVM, and VirtualBox多种虚拟化技术。

在这种某一台机器的虚拟化的环境中，一个虚拟交换机（vswitch）主要有两个作用：传递虚拟机VM之间的流量，以及实现VM和外界网络的通信。

在openstack中目前用的比较多的L2层agent应该就是openvswitch agent了。本文大致分析了一下openvswithc agent做了哪些事。

看一下openvswitch agent的启动：

1.neutron/plugins/openvswitch/agent/ovs_neutron_agent.py:main() 2. plugin = OVSNeutronAgent(**agent_config) 3. self.setup_rpc() 4. self.plugin_rpc = OVSPluginApi(topics.PLUGIN) 5. self.state_rpc = agent_rpc.PluginReportStateAPI(topics.PLUGIN) 6. self.connection = agent_rpc.create_consumers(…) 7. heartbeat = loopingcall.FixedIntervalLoopingCall(self._report_state) 8. self.setup_integration_br() 9. self.setup_physical_bridges(bridge_mappings) 10. self.sg_agent = OVSSecurityGroupAgent(…) 11. plugin.daemon_loop() 12. self.rpc_loop() 13. port_info = self.update_ports(ports) 14. sync = self.process_network_ports(port_info) 复制代码

启动时做了以下工作：

1. 设置plugin_rpc，这是用来与neutron-server通信的。 2. 设置state_rpc，用于agent状态信息上报。 3. 设置connection，用于接收neutron-server的消息。 4. 启动状态周期上报。 5. 设置br-int。 6. 设置bridge_mapping对应的网桥。 7. 初始化sg_agent，用于处理security group。 8. 周期检测br-int上的端口变化，调用process_network_ports处理添加/删除端口。 首先boot虚机时，nova-compute发消息给neutron-server请求创建port。之后，在driver里面在br-int上建立port后，neutron-openvswitch-port循环检测br-int会发现新增端口,对其设定合适的openflow规则以及localvlan，最后将port状态设置为ACTIVE。

所以我们要做的是把vm3 上的neutron-openvswitch-agent 给变成正常，

命令：/sbin/service openstack-neutron-openvswitch-agent restart