Shulou(Shulou.com)06/03 Report--

1. Brief introduction of environment

1.1 hardware environment

Prepare 4 physical servers as deployment nodes

Go to the host BIOS and turn on the Intel (R) Virtualization Technology and Intel (R) VT forDirected Icano virtualization support features.

The official requirements for the oVirt hardware environment are shown in the figure.

1.2 Software environment

All server systems: CentOS7.3 minimal 64-bit

OVirt version: 4.1.1

2. Introduction to oVirt

OVirt is a management tool for building a private cloud platform based on KVM virtualization. OVirt manages the KVM virtualized environment in the way of web interface, which is mainly composed of Engine and Node. Engine is the management node in oVirt, which provides web interface to interact with users and makes Node node management visual. Node is the computing node in oVirt. The underlying layer focuses on KVM virtualization and provides computing resources for virtual machine operation. OVirt is the solution of data center virtualization management platform, and data center is the logical concept, divided according to the organization of master data storage, in order to quickly build private cloud, mainly used in desktop cloud and server virtualization.

OVirt Engine is responsible for the management and maintenance of the whole system is the core component of the entire oVirt management platform. The Engine node built-in database records all the virtual machine configuration, node status, system environment, memory and other information in the whole system. Realize the logic, state and policy of management, and a good web interface client, so that the system administrator can manage any operation of the system through the web page. OVirt Node is only responsible for providing computing resources and functional implementation, does not record any information, and is only responsible for the creation and modification of virtual machines and devices, and the sharing and protection of resources. The Node node can be made up of VDSM components installed in an ordinary Linux operating system, or it can be a Linux operating system customized for Node. VDSM is the link that maintains the relationship between Node nodes and Engine. It is responsible for the self-startup and registration of nodes and the management of networks, storage, hosts and virtual machines. It supports the external interference of virtual machines, the merging and overspending of memory and storage and other functions. The content of the Node node system disappears after the restart, thus ensuring the statelessness of the Node.

OVirt official website: http://www.ovirt.org/

OVirt official website deployment manual:

Http://www.ovirt.org/documentation/install-guide/Installation_Guide/

3. Deployment practice

According to the requirements of the oVirt schema architecture diagram, three servers are used to deploy the oVirt Node compute node and one server to deploy the oVirt Engine management node. OVirt can be installed either in a virtual machine or on a physical server. This installation uses a physical server, and virtual machine installation is not recommended. When installing with a virtual machine, the running virtual machine cannot enter the console page after installation.

3.1 Environmental preparation

Deploy the underlying system

OVirt Virtualization all services require the Linux system as the underlying running environment. Therefore, install the CentOS7.3 minimal 64-bit operating system on all four physical servers. Configure the correct time and time zone (Shanghai).

Turn off all host firewalls:

# systemctl stop firewalld.service

# systemctl disable firewalld.service

Modify the hostname:

Take engine as an example, as follows:

# hostnamectl-static set-hostname engine.taocloud.com

Configure the remaining Node node hostnames.

Configure yum Feed

OVirt-Engine uses the installation method of yum source and configures the oVirt yum source of all hosts. The download speed of the yum source provided on the official website is too slow, so it is not recommended. Execute the following commands on each server:

Sudo yum install http://mirror.isoc.org.il/pub/ovirt/yum-repo/ovirt-release41.rpm

Engine nodes need yum to install oVirt-Engine,Node nodes need to deploy VDSM and KVM services.

3.2Engine management node

Install the management node

As a management node, Engine is the core component of the whole system, which controls the use of computing node resources and provides a web management interface for users to access.

After all the yum sources are configured, install the oVirt-Engine management node and execute the following command:

Sudo yum install ovirt-engine-y

Configure oVirt Engine

Execute command: ovirt-engine setup

Except for some of the names that need to be filled in, all of them are default, and the configuration information is shown in the following figure.

Log in to the Engine Web administration interface

Ovirt Engine uses FQDN for web access. If DNS is not configured, you need to modify the hosts file of the remote Windows machine, as shown below.

Open a browser on the remote windows side and enter: http://ovirtengine.taocloud.com/ovirt-engine to enter the oVirt WEB interface. Log in using the administrative portal.

If you do not modify the hosts file to resolve the host name, you will not be able to access the WEB interface with the normal name.

Use the default data center of the system and choose shared storage when configuring the cluster.

3.3Node compute node

Ovirt Node supports two deployment methods. VDSM and KVM services can be installed in ordinary Linux systems to build Node nodes, and oVirt Node customized Linux operating system can also be installed between servers. This deployment adopts the method of deploying services by ordinary Linux system.

Install the compute node

As the computing node of the virtualized environment, oVirt Node is responsible for providing computing resources. When building, VDSM and KVM services need to be deployed in the Linux operating system. Before installing the service, you need to ensure that the server has the following conditions:

Can ping access to the extranet

The yum source is configured correctly

Execute the command: # setenforce 0 # close SElinux

# vi / etc/selinux/config # modify the configuration file to ensure that the status remains unchanged after restart

# SELINUX = enforcing # comment out the line change and add information under it

SELINUX=permissive

After saving and exiting. Execute the command: # getenforce

The result returned by Permissive # indicates that the configuration is correct

After the above conditions are met, go back to the Engine management interface and add the oVirt Node node directly through ip. You can use ip to write the name at will. Be sure to enter the correct node ip and verification password, and fill in the rest according to your needs. After waiting for a period of time, the oVirt Node installation is completed to ensure that there may be warnings but no errors during the installation process when the above conditions are met.

Node Custom Image

Download Node custom Linux system image from ovirt official website:

Ovirt-node-ng-installer-ovirt-4.1-2017041904.iso

Deploy the system image directly on the Node server, and then add the Node node in the Engine management interface.

3.4 Storage domain

Configure nfs Stora

OVirt cluster has two methods: local storage and shared storage. When local storage is selected, only a single Node node can be added to the cluster. Therefore, in order to make better use of the cluster characteristics, nfs shared storage is adopted. Configure nfs shared storage on each Node node to install virtual machines, and execute the following command:

Mkdir / export/iso # creates an iso storage domain to hold system images

Mkdir / export/data # creates a data storage domain for storing virtual machine data

Chown 36:36-R / export # Grant VDSM permission

Chmod 0755-R / export # gives executive permission

Cat / etc/exports # Edit the exports file as follows

/ export/ iso * (rw,sync,no_subtree_check,all_squash,anonuid=36,anongid=36)

/ export/ data* (rw,sync,no_subtree_check,all_squash,anonuid=36,anongid=36)

Systemctl start nfs.service # start the nfs service

Systemctl enable nfs.service # set the nfs service to boot automatically

Add NFS Stora

Create a new oVirt storage domain to store data in the data storage domain management interface, and then create a new ISO storage domain to store virtual machine images, and fill in the correct storage path and other information according to the page prompts.

Upload image

After adding iso storage domain and data data storage domain in the interface, you can copy the system image directly to the directory / export/ iso/98e1a52b-b72f-40f5-8c3e-f1cbb8989b93/p_w_picpaths/11111111-1111-1111-1111-111111111111. After the copy is completed, you can refresh it in the foreground interface and use it.

3.5 Virtual Machine

Create a virtual machine

When creating a virtual machine, you must at least attach or create an instance image, that is, a virtual machine system disk, add at least one network, and click OK after editing the specification requirements.

Select the virtual machine in the virtual machine list interface, click "run only once", enter the boot option to select the appropriate system image, and then open the console to install the operating system.

4. common problem

Error 1:

If there are any errors in the execution of yum install ovirt-engine, check whether there is a problem with the yum source. Generally, the configuration of the yum source is correct and there will be no errors. If you are unsure, try to replace the Yum source.

Error 2:

Error adding Node node to the administrative interface:

Host installation failed. Command returnedfailure code 1 during SSH session 'root@192.168.*.*'

Resolve:

(1) check whether the ovirt sources in / etc/yum.repos.d are complete. You must have ovirt-4.1-dependencies.repo and ovirt-4.1.repo, and the following CentOS-*, is the normal source.

[root@node01 ~] # ls / etc/yum.repos.d/

Bak CentOS-Debuginfo.repo CentOS-Sources.repo ovirt-4.1.repo

CentOS-Base.repo CentOS-fasttrack.repo CentOS-Vault.repo server.repo

CentOS-CR.repo CentOS-Media.repo ovirt-4.1-dependencies.repo

Yum clean all can be performed to clean up the yum source cache.

(2) check to see if you can ping the extranet.

(3) / etc/init.d/rpcbind restart / etc/init.d/vdsmd restart

(4) check the status of firewalld and selinux. If the addition still fails, you can try to execute it:

You can follow this method to troubleshoot any failure to add a Node node.

Error 3:

If you fail to mount nfs, go to the underlying layer and test it with instructions:

[root@node1] # showmount-e 192.168.3.233

Rpc mount export: RPC: Unable to receive;errno = No route to host

Close firewalld and iptables

Attachment: http://down.51cto.com/data/2366737

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.