Based on the needs of the simulation environment, after measuring the many existing open source tools, we decided to choose OpenStack as the experimental environment.
The first step is to deploy the OpenStack environment, and then consider the way OpenStack is deployed. There are currently three main categories:
1. Compile the installation with source code on Linux.
2. Install using the binary installation package on Linux.
3. Install using the secondary installation tool (Fuel,openshit, etc.).
The difficulty level from 1 to 3 is high, medium, and low. For beginners like me, I recommend scenario 3. The first two installation process is too complicated, the installation process may also need to install a lot of other dependencies, and error frequency. It is easy to dispel the enthusiasm of beginners, so the first-class students recommend using the first to understand OpenStack, familiar with the interest can be used to install.
Here, I give a link to the Fuel-openstack experience, which is simply an experience, and there are not too many features to work with.
Http://demo.fuel-infra.org:8000/#clusters
Where the user name and password are admin.
Then about the installation of 3, there are many tutorials online, here do not do too much explanation. Give only a few detailed links for your reference.
Http://www.cnblogs.com/dyllove98/p/4115629.html
Http://www.cnblogs.com/heybob/p/4718927.html
The above two have illustrated the installation process of fuel, there is nothing to say. The main thing I want to express is one of the small problems, that is why the URL given by fuel server cannot access the fuel UI.
The problem itself is related to the planning of the network address, and both VirtualBox and VMware have their own set of propositions. Make a brief description of each:
VirtualBox:
The above two links are defined in VBox under the three Global Virtual network card (host-only mode): 10.20.0.1;172.16.0.1;192.168.4.1, respectively, corresponding to a, B, C class address. It is also three different network segments, which are responsible for OpenStack internal communication, floating IP, data service and so on. Consider the design of these addresses for the moment, but it must be understood that 10.20.0.1 this address is the default IP address of the fuel node, so it is recommended that this address remain the same.
Then, after I've set all the tutorials, I can't access the fuel UI. After flipping through the VBox ref, the VNet0 in the VBox is modified to be the IP of the same network segment as the host (the three networks above are the same, vnet1,2,3) and can be accessed instantaneously. The analysis is as follows:
For the VBox host-only mode, the host and the Virtual machine Communication network card is VNET0, and the default Net0 and the host's IP is not in the same network segment, so can not communicate, modified, then it is done. VMware is the same principle (although I did not install it with VMware).
Vmware:
For VMware's use, it has three virtual network cards by default: VMnet0, VMnet1, VMnet8, where 0 corresponds to bridge mode, 1 corresponds to host-only mode, 8 corresponds to NAT mode, and communication channels for virtual machines and hosts. Therefore the self-created network segment should not use these three. I have probably painted this structure as shown in:
The above concept is only for individuals in the process of groping installation fuel experience, there may be a misunderstanding of the place, welcome correction. Also in the online search fuel UI access solution, found a lot of people have this problem, but no one answer, therefore have this blog post, purely original, I hope reprint annotated source.
Network address planning and communication mechanisms for installing fuel-openstack using VirtualBox and VMware