www.JeBas.us - planning

Building a Home Data Center for Fun and Profit.

JeBas — Mon, 23 Feb 2009 10:00:00 +0000

A while back, the gentoo-wiki and gentoo-portage websites got yanked off of the net because the ISP got into a dispute with its landlord. This got me wondering if it was possible for a small business owner or hobbyist to build their own data center. So I am starting a project to do just that.

The first step for me is to move my existing servers off of the physical hardware and onto virtual servers. Virtual servers can easily be moved from one physical system to another. As a bonus it is possible to clone a virtual machine to test upgrades before they go into production. Also, none of my present servers use enough I/O, CPU, or RAM to require an independent box.

After a little research, I chose my virtualization software based on my requirements.

I eliminated OpenVZ fairly early. Though most of my servers are Linux based, I occasionally need access to a Windows machine. OpenVZ uses one main kernel to run all of the separate environments, and will not run Windows.

I also eliminated Virtual Box and QEMU. These system seemed to be designed to run on desktops environments. Both have heavy emphasis on the user interface. Neither of them seem to be interested in server cloning or clusters.

KVM was eliminated because of my hardware. KVM requires the Intel VT or AMD-V, and some of my machines don't have it. Since I don't have the money to replace these machines, I can't use KVM.

VMWare was eliminated for the same reason that KVM was. Though VMWare will run on machines without Intel VT or AMD-V, it will not run a 64 bit OS. VMWare is limited to using 32 bit OSes if it does not have the hardware support.

This left me with Xen. It will run 64 bit OS as long as that OS is built for paravirtualization. It will limit which machines will run Windows, but I don't see that as a problem. It supports server cloning, and allows servers to be moved to different physical machines.

So my first step is to move JeBas.us off of its present server, and onto a Xen virtual machine.

Why I chose Drupal over Joomla.

JeBas — Mon, 15 Dec 2008 10:00:00 +0000

When I originally started using this domain, I was using Joomla. I wanted the experience of running a website, but I also knew that I wanted to do a few things more than were available in WordPress at the time. So I loaded Joomla onto my Apache server, and started my stories blog.

As I got use to putting up posts, I quickly realized that there were other subjects that I wanted to write about. Some of these subjects required more than a simple content management system. I had wanted to have recipes on my GlutenFree site and galleries on my art site. I also wanted to link all of my sites so that there was a common login for everyone. That's when I realized that Joomla was not going to work for me.
The first problem that I ran into was how modules were distributed in Joomla. Most of the modules that I was interested in were proprietary, and shipped without the source code. I didn't have a problem with paying someone for providing the module, but a lot of what I was interested in doing required changes to the modules. This made modifying the site more difficult.

The second problem was running multiple sites off of the same code. Joomla, at the time, was designed to run one site at a time. My goal at the time was to have multiple sub-domains sites running off a common login database. There was a module I could added that was highly unstable and extremely difficult to install. It would give me multiple site access, but it would have made software upgrades difficult. I was forced to return to my original research, and choose some new software.
When I had done my original research, I was deciding between Drupal, Joomla, and Plone. I had given up on Plone because I could never get it to run at a reasonable speed, so I looked at Drupal again.

I immediately noticed that Drupal placed its modules into an open source format. If something didn't quite meet my needs, I would have access to the code to make changes. Secondly, the system was initially designed to run multiple sites off the same code and database. There was even a common sign on module that gave me the ability to have a single login for all of the sites.

After a little time for debate, I finally decided that it was worth scrapping Joomla, and rebuilding the site with Drupal. It was a bit of a pain, but Drupal has proven that it was worth the effort.

New Plan for www.JeBas.US

JeBas — Mon, 17 Nov 2008 10:00:00 +0000

In the past year, I have learned a lot developing JeBas.US. Some of the lessons were technical, some were spiritual, and some were just plan weird. However the most important lesson that I've learned is that you need to contribute as much if not more than you receive. Because of that, I am now changing the www.JeBas.US website.

All of these various lessons that learned are now in little notes all over my office. Most of them could be useful to other people if I were to post them somewhere. This includes experimentation with virtual machines, comparisons of Drupal and Joomla, and what makes a working website. It's a lot of information that I am now going to share.

I'm going to be discussing the various aspects of running this website. That means the topics may be a little broad, but they will have a narrow focus. It will cover what I have done, some of the mistakes that I've made, and some of my successes. It will also cover sections as I learn a new skill for this on going project.

I am hoping that the information that I am providing will be as useful to someone else as I have found it.