OpenStack问题集锦（持续更新中...）_libvirt disk cache mode

如何再添加一个管理员用户

# keystone-manage bootstrap --bootstrap-username myadmin --bootstrap-password mypasswd

1.查看某个实例的url（例如访问前面创建的名为instance-a 的实例）

$ openstack console url show 5b08017b-00d4-4476-9380-4f5b6165c6d7

+-------+---------------------------------------------------------------------------------+

| Field | Value                                                                           |

+-------+---------------------------------------------------------------------------------+

| type  | novnc                                                                           |

| url   | http://controller:6080/vnc_auto.html?token=6643713d-f4c8-411c-ac9e-2c5b5a419935 |

+-------+---------------------------------------------------------------------------------+

（5b08017b-00d4-4476-9380-4f5b6165c6d7 通过openstack server list 查看，是要启动实例的ID）

2.访问一个实例

（在浏览器中输入http://controller:6080/vnc_auto.html?token=6643713d-f4c8-411c-ac9e-2c5b5a419935 就可以访问虚拟机了，前提是controller主机名能被识别，或直接换成ip）

如何设置虚拟机磁盘类型，如ide、virtio

创建虚拟机时直接传入"disk_bus": "ide" ：

 

Action: 'create', calling method: <bound method ServersController.create of <nova.api.openstack.compute.servers.ServersController object at 0x7f1e5c008510>>, body: {"server": {"name": "test-1", "imageRef": "", "availability_zone": "nova", "block_device_mapping_v2": [{"disk_bus": "ide", "source_type": "image", "boot_index": "0", "uuid": "907bcb20-7623-446f-b19c-ab135a0ac2d6", "destination_type": "volume", "volume_size": 400, "delete_on_termination": true, "disk_name": "test-1-disk1", "device_name": "vda"}], "flavorRef": "8a6558a8-2d36-48d9-bd0d-2fbe1ac25a2e", "max_count": 1, "interface_disconnect_names": [], "min_count": 1, "networks": "none"}}

如何设置虚拟机的cpu拓扑

$ openstack flavor set FLAVOR-NAME \

    --property hw:cpu_sockets=FLAVOR-SOCKETS \

    --property hw:cpu_cores=FLAVOR-CORES \

    --property hw:cpu_threads=FLAVOR-THREADS \

    --property hw:cpu_max_sockets=FLAVOR-SOCKETS \

    --property hw:cpu_max_cores=FLAVOR-CORES \

    --property hw:cpu_max_threads=FLAVOR-THREADS

 

或者

$ nova flavor-key m2.nano set hw:cpu_sockets=3  hw:cpu_cores=4  hw:cpu_threads=5

如何设置虚拟机的操作系统类型

# openstack image set --property os_type=windows 950420e3-98ae-454b-b073-9f75b275fcc8

或者

# nova image-meta 950420e3-98ae-454b-b073-9f75b275fcc8 set os_type=linux

 

这样在镜像的属性里面就能看到os_type

如何设置虚拟机磁盘的cache类型

修改配置文件中的disk_cachemodes：

/etc/nova/nova-compute.conf

[libvirt]

disk_cachemodes = "network=writeback"

 

其中network和writeback分别对应虚拟机xml中的type和cache：

  <devices>

    <disk type=' network' device='disk'>

      <driver name='qemu' type='raw' cache=' writeback'/>

这个配置是不会写入数据库的，每次虚拟机启动时根据这个配置文件动态生成。

如何设置虚拟机cpu类型

修改配置文件中的cpu_mode:

/etc/nova/nova-compute.conf

[libvirt]

cpu_mode=none

这个配置是不会写入数据库的，每次虚拟机启动时根据这个配置文件动态生成。

如果cpu_mode=none:

  <cpu>

      <topology sockets='1' cores='1' threads='1'/>                           

   </cpu>

如果cpu_mode=host-model:

  <cpu mode='host-model'>

      <model fallback='allow'/>

      <topology sockets='1' cores='1' threads='1'/>                           

   </cpu>

如何设置虚拟机cpu的亲和性

修改/etc/nova/nova.conf

[DEFAULT]

vcpu_pin_set = 0-1

使用上面的配置后虚拟机的xml文件如下：

  <vcpu placement='static' cpuset='0-1'>2</vcpu>

其作用主要是让openstack的虚拟机使用指定的几个物理cpu核，给物理机操作系统适当留下若干物理核，保证物理机的性能。

说明：另外openstack还可以通过flavor隐式的做虚拟机cpu的绑定，通过做host的aggregate：

The flavor extra specs will be enhanced to support twonew parameters

hw:cpu_policy=shared|dedicated

hw:cpu_threads_policy=avoid|separate|isolate|prefer

如何设置虚拟机的关联组与反关联组

nova server-group-create group_name anti-afffinity

nova boot --image cirros --flavor m1.small --hint group=group_name

如何更新网络的ip地址池的范围

neutron subnet-update --allocation-pool start=192.168.105.35,end=192.168.105.39 --allocation-pool start=192.168.105.40,end=192.168.105.49 subnet_01

如何使用floating ip

neutron floatingip-create providernet01

neutron floatingip-associate  fbc76a22-4a2e-43bb-982b-6e42575eebe9 12b50cb3-18c8-4c08-8caf-1203e3183406

(floating_id   port_id)

如何在subnet中创建port

neutron port-create --fixed-ip subnet_id=$SUBNET_ID $NET_ID

如何控制磁盘qos（cinder）

一。首先创建一个带限速功能的类型

# cinder type-create my-type 

# cinder qos-create my-qos consumer="front-end" read_iops_sec=2000 write_iops_sec=1000

# cinder qos-associate $QOS_ID $VOLUME_TYPE_ID

# cinder qos-get-association $QOS_ID

 

二。把创建的类型赋给磁盘

如果是已经存在的磁盘：

1.从虚拟机分离磁盘

2.#  cinder retype $VOLUME_ID  $VOLUME_TYPE_ID

3.给虚拟机挂载磁盘

 

2.给虚拟机挂载磁盘

说明：

      <iotune>

        <read_iops_sec>2000</read_iops_sec>

        <write_iops_sec>1000</write_iops_sec>

      </iotune>

如何控制磁盘qos（nova）

# openstack flavor create  --vcpus 1 --ram 4096 --disk 5 qqq

# openstack flavor set qqq --property quota:disk_read_bytes_sec=20480 --property quota:disk_write_bytes_sec=10240

    <disk type='network' device='disk'>

      <driver name='qemu' type='raw' cache='writeback' discard='unmap'/>

      <source protocol='rbd' name='images/1a956ba7-25fe-49f1-9513-7adb8928036c'>

        <host name='192.168.107.50' port='6789'/>

        <host name='192.168.107.51' port='6789'/>

        <host name='192.168.107.52' port='6789'/>

        <host name='192.168.107.53' port='6789'/>

      </source>

      <target dev='vda' bus='virtio'/>

      <iotune>

        <read_bytes_sec>20480</read_bytes_sec>

        <write_bytes_sec>10240</write_bytes_sec>

      </iotune>

      <boot order='1'/>

      <address type='pci' domain='0x0000' bus='0x00' slot='0x0a' function='0x0'/>

    </disk>

# openstack flavor create  --vcpus 1 --ram 4096 --disk 5 qqqq

# openstack flavor set qqqq --property quota:disk_read_iops_sec=20480 --property quota:disk_write_iops_sec=10240

    <disk type='network' device='disk'>

      <driver name='qemu' type='raw' cache='writeback' discard='unmap'/>

      <source protocol='rbd' name='images/1a956ba7-25fe-49f1-9513-7adb8928036c'>

        <host name='192.168.107.50' port='6789'/>

        <host name='192.168.107.51' port='6789'/>

        <host name='192.168.107.52' port='6789'/>

        <host name='192.168.107.53' port='6789'/>

      </source>

      <target dev='vda' bus='virtio'/>

      <iotune>

        <read_iops_sec>20480</read_iops_sec>

        <write_iops_sec>10240</write_iops_sec>

      </iotune>

      <boot order='1'/>

      <address type='pci' domain='0x0000' bus='0x00' slot='0x0a' function='0x0'/>

    </disk>

如何控制网络qos（neutorn）

 

如何控制网络qos（nova）

# openstack flavor create  --vcpus 1 --ram 512 --disk 1 q

# openstack flavor set q --property quota:vif_inbound_average=2048  --property quota:vif_outbound_average=1024

    <interface type='bridge'>

      <mac address='fa:16:3e:8f:6a:c9'/>

      <source bridge='brq5233ef6c-62'/>

       <bandwidth>

        <inbound average='2048'/>

        <outbound average='1024'/>

      </bandwidth>

      <target dev='tap9573c869-24'/>

      <model type='virtio'/>

      <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/>

    </interface>

如何在数据库中删除虚拟机

MariaDB [nova]> delete from instance_actions_events where action_id="129";

MariaDB [nova]> delete from instance_actions_events where action_id="138";

 

MariaDB [nova]> delete from instance_actions where instance_uuid="67b6044b-bbdc-4120-b2c7-c6faf540da29"; （id和action_id关联）

 

MariaDB [nova]> delete from block_device_mapping where instance_uuid="67b6044b-bbdc-4120-b2c7-c6faf540da29";

 

MariaDB [nova]> delete from instance_extra where instance_uuid="67b6044b-bbdc-4120-b2c7-c6faf540da29";

 

MariaDB [nova]> delete from instance_info_caches where instance_uuid="67b6044b-bbdc-4120-b2c7-c6faf540da29";     

 

MariaDB [nova]> delete from instance_system_metadata where instance_uuid="67b6044b-bbdc-4120-b2c7-c6faf540da29";

 

MariaDB [nova]> delete from instances where display_name="test-1";