OpenStack Computing Service Nova [IV] 03/31 Update SLTechnology News&Howtos

OpenStack Computing Service Nova [IV]

2025-03-31 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Database >

Shulou(Shulou.com)06/01 Report--

OpenStack Computing Service Nova [IV] OpenStack Computing Service Nova [IV]

Openstack

Time: November 28, 2016

Introduction to Nova:

Nova is one of the first two modules of openstack, and the other is object storage swift. In the openstack system, one is called a computing node, and the other is called a control node. This is mainly related to nova. We install it as the compute node nova-compute and call it the control node except nova-compute. Nova-compute creates a virtual machine, just creates a virtual machine, and all the control is on another machine.

Nova has a lot of components (services)

Introduction to Nova Service

API: responsible for accepting and responding to external requests. Support for OpenStack API,EC2 API

Cert: responsible for authentication EC 2

Scheduler: used for CVM scheduling

Conductor: middleware for computing nodes to access data

Consoleauth: authorization verification for the console

Novncproxy:VNC Agent

Hint: EC 2 means Amazon CVM.

Nova scheduler

The role of the Nova scheduler module in openstack is to decide which host (compute node) the virtual machine is created on.

There are two steps to decide which physical node a virtual machine should be scheduled to:

1. Filter (Fliter)

2. Calculate the weight (Weight)

Tip: we often can not find a valid host? Why?

Because nova scheduler thinks you don't have the resources to create a virtual machine, even if you have 100 gigabytes of memory, you can't create it if nova scheduler thinks you're not qualified. The function of scheduler is to decide which host the virtual machine is created on.

After being filtered by the host, you need to calculate the weight of the host and select a corresponding host according to the policy (for each virtual machine to be created)

Control node settings:

We have already done the database and keystone modification earlier, so skip it here

Install the package

[root@linux-node1 ~] # yum install openstack-nova-api openstack-nova-conductor\ > openstack-nova-console openstack-nova-novncproxy\ > openstack-nova-scheduler

Modify the connection address of the database in the configuration file

[root@linux-node1] # vim / etc/nova/nova.conf... [database] connection=mysql+pymysql://nova:nova@192.168.56.11/nova... [api_database] connection=mysql+pymysql://nova:nova@192.168.56.11/nova_api

Tip: don't make any mistakes. Each one is modified under the corresponding module.

Synchronize database

[root@linux-node1] # su-s / bin/sh-c "nova-manage api_db sync" nova [root@linux-node1 ~] # su-s / bin/sh-c "nova-manage db sync" nova

Hint: db's warning can be ignored.

Check if there is a table structure

[root@linux-node1 ~] # mysql-h 192.168.56.11-unova-pnova-e "use nova;show tables;" [root@linux-node1 ~] # mysql-h 192.168.56.11-unova_api-pnova_api-e "use nova_api;show tables;"

Configure keystone

[root@linux-node1] # vim / etc/nova/nova.conf... [keystone_authtoken] auth_uri = http://192.168.56.11:5000auth_url = http://192.168.56.11:35357memcached_servers = 192.168.56.11:11211auth_type = passwordproject_domain_name = defaultuser_domain_name = defaultproject_name = serviceusername = novapassword = nova... [DEFAULT] auth_strategy=keystone# opens comments, and this line represents authentication using keystone

RabbitMq (message queuing configuration)

Because nova services communicate with each other using message queuing, we need to configure rabbitmq

[root@linux-node1 ~] # vim / etc/nova/ nova.confession [default] rpc_backend=rabbit# search rpc_backend Open comment rabbit_host=192.168.56.11 modify localhost to IP address rabbit_port=5672 port we can open or not open the default rabbit_userid=openstackrabbit_password=openstack this is the user we configured on rabbitmq

Configure some of nova's own features

Set the enabled metadata API [root@linux-node1 ~] # vim / etc/nova/ Nova.confs [default] enabled_apis=osapi_compute,metadata to start the network service support use_neutron=true to turn off the firewall firewall_driver=nova.virt.firewall.NoopFirewallDriver

Tip: by default, the computing service uses the built-in firewall service. Because network services include firewall services, you must use nova.virt.firewall.NoopFirewallDriver firewall services to disable the firewall services built into computing services

We do not configure my IP because My IP is a big hole

Configure the VNC agent to use the management interface IP address of the control node

[root@linux-node1 ~] # vim / etc/nova/nova.confvncserver_listen=192.168.56.11vncserver_proxyclient_address=192.168.56.11

Configure glance mirroring service API

[root@linux-node1 ~] # vim / etc/nova/ Nova. Confession [glance] api_servers= http://192.168.56.11:9292

Configure lock path

[oslo_concurrency]... Lock_path=/var/lib/nova/tmp

Nova configuration description

[root@linux-node1 ~] # grep'^ [a Murz]'/ etc/nova/nova.confenabled_apis=osapi_compute Metadata # start apiauth_strategy=keystone # set up keystonefirewall_driver=nova.virt.firewall.NoopFirewallDriver # turn off firewall use_neutron=true # use neutronrpc_backend=rabbit # use rabbitmqconnection = mysql+pymysql://nova:nova@192.168.56.11/nova_api # database address Connection = mysql+pymysql://nova:nova@192.168.56.11/nova # database address api_servers= http://192.168.56.11:9292 # glance api address auth_uri = http://192.168.56.11:5000 # keystoneauth_url = http://192.168.56.11:35357 # keystonememcached_servers = 192.168.56.11 http://192.168.56.11:35357 11211 # keystoneauth_type = password # keystoneproject_domain_name = default # keystoneuser_domain_name = default # keystoneproject_name = service # keystone username = nova # keystonepassword = nova # keystonelock_path=/var/lib/nova/tmp # Lock path rabbit_host=192.168.56.11 # rabbitmqrabbit_port=5672 # rabbitmqrabbit_userid=openstack # rabbitmqrabbit_password=openstack # rabbitmqvncserver_listen=192.168.56.11 # VNCvncserver_proxyclient_address=192.168.56.11 # VNC

Set up boot and start the service

# systemctl enable openstack-nova-api.service\ openstack-nova-consoleauth.service openstack-nova-scheduler.service\ openstack-nova-conductor.service openstack-nova-novncproxy.service# systemctl start openstack-nova-api.service\ openstack-nova-consoleauth.service openstack-nova-scheduler.service\ openstack-nova-conductor.service openstack-nova-novncproxy.service

After nova starts successfully, you still need to register on keystone, otherwise others will not be able to connect

Create a nova service

Create a Compute service api endpoint

Check whether the control node is successful

How many virtual machines we want to build in the architecture depends on the configuration of computing nodes. We can only use VMware for virtualization tools, because VMware supports nested virtualization, while other virtualization software does not.

Our nova computing node IP is 192.168.56.12

Virtualization technology needs to be turned on

Because we need to use kvm to create virtual machines, we need to turn on virtualization. If the server needs to be enabled on bios

Environmental preparation

[root@linux-node2 ~] # cat / etc/redhat-release CentOS Linux release 7.2.1511 (Core) [root@linux-node2 ~] # uname-r3.10.0-327.36.2.el7.x86_64

Time synchronization

[root@linux-node1 ~] # yum install ntpdate-y [root@linux-node1 ~] # ntpdate time1.aliyun.com [root@linux-node1 ~] # timedatectl set-timezone Asia/Shanghai # set time zone [root@linux-node1 ~] # rpm-ivh http://mirrors.aliyun.com/epel/epel-release-latest-7.noarch.rpm

Install the openstack repository

[root@linux-node2 ~] # yum install-y centos-release-openstack-mitaka

Install the openstack client

[root@linux-node1 ~] # yum install-y python-openstackclient

Because the nova of the control node and the configuration of the compute node are the same except that there is no database, here we modify it in the form of scp

Install the openstack SELinux Management Pack

[root@linux-node2 ~] # yum install-y openstack-selinux

Install nova

[root@linux-node2 ~] # yum install-y openstack-nova-compute

Steps:

1. Scp nova.conf from the control node

2. Delete the configuration of database

3. Change the configuration of vnc

4. Set an option for virtualization

Tip: pay attention to the permissions of the nova.conf file

Time must be synchronized!

1. Copy the nova.conf of the control node to the compute node

[root@linux-node1] # scp / etc/nova/nova.conf 192.168.56.12:/etc/nova/The authenticity of host '192.168.56.12 (192.168.56.12)' can't be established.ECDSA key fingerprint is 43:50:3c:fa:03:29:7c:3c:5f:aa:d2:76:b5:8e:d9:54.Are you sure you want to continue connecting (yes/no)? YesWarning: Permanently added '192.168.56.12' (ECDSA) to the list of known hosts.root@192.168.56.12's password: nova.conf 100% 180KB 180.0KB/s 00:00

two。 Enter the computing node to modify

First of all, make sure the permissions are consistent [root@linux-node2 nova] # ll / etc/nova/nova.conf-rw-r- 1 root nova 184332 Nov 18 17:02 / etc/nova/nova.conf

3. Modify the configuration file

[root@linux-node2 nova] # vim / etc/nova/nova.conf#connection = # connection = search mysql to comment out the mysql path

Configure vnc

Novncproxy_base_url= http://192.168.56.11:6080/vnc_auto.htmlvncserver_listen=0.0.0.0vncserver_proxyclient_address=192.168.56.125384 Line enabled=true

Tip: the server component listens for all IP addresses, while the proxy component listens for only the IP address of the compute node's management network interface. The basic URL indicates where you can access the remote console of the instance on that compute node using a web browser.

Select Virtualization Type

Determine if your compute node supports virtualized hardware acceleration.

Egrep-c'(vmx | svm)'/ proc/cpuinfo

If this command returns a value of 1 or not equal to 0, then your compute node supports hardware acceleration and does not require additional configuration.

If this command returns a value of 0, then your compute node does not support hardware acceleration. You must configure libvirt to use QEMU instead of KVM

KVM article: http://www.abcdocker.com/abcdocker/1627

Make the following edits in the [libvirt] area of the / etc/nova/nova.conf file:

[libvirt]... Virt_type=kvm# configuration Virtualization Type

Summary

Nova.conf modified the following five lines

[root@linux-node2 nova] # grep'^ [a Murz]'/ etc/nova/nova.conf... Enabled=truevncserver_listen=0.0.0.0vncserver_proxyclient_address=192.168.56.12novncproxy_base_url= http://192.168.56.11:6080/vnc_auto.htmlvirt_type=kvm

Set up boot boot

[root@linux-node2 ~] # systemctl enable libvirtd openstack-nova-compute [root@linux-node2 ~] # systemctl start libvirtd openstack-nova-compute

List the service components to verify that each process was successfully started and registered:

Check to see if nova and keystone are normal

[root@linux-node1 ~] # nova service-list+----+-+--+-- -+ | 1 | nova-consoleauth | linux-node1.abcdocker.com | internal | enabled | up | 2016-11-18T09:24:23.000000 |-| | 2 | nova-conductor | | linux-node1.abcdocker.com | internal | enabled | up | 2016-11-18T09:24:22.000000 |-| | 3 | nova-scheduler | linux-node1.abcdocker.com | internal | enabled | up | 2016-11-18T09:24:23.000000 |-| 6 | nova-compute | linux-node2.abcdocker.com | nova | enabled | up | 2016-11-18T09:24:23.000000 |-| | | +-+ -+

Check whether nova and glance services are normal with each other

Hint again: time must be synchronized!

This is the end of nova installation!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.