Even Virtual PC Can Crash

I run a lot of instances of Microsoft's Virtual PC to keep my personal and business records separate from my development workstation. I find that this separation very convenient where I the virtual disk drive with the OS and application data is small enough to fit on a single DVD.

This way the Virtual PC's are very portable where I can take the DVD and copy it to a new workstation and continue work in the event that I need to upgrade to a new workstation. Now reinstallation, no mess, just takes a couple of minutes to copy the file to the new machine.

But Virtual PC's do crash.

The typical scenario that you will run into is when you have automatic updates enabled or install new applications that modify the OS or drivers.

My last virtual crash happened when I installed this year's TurboTax. The new version required .NET 2.0 so it had to download upgrade from .NET 1.0. Remember, I keep these virtual drives small enough to fit on a DVD, so the install failed and left the drive with zero space.

I run the disk cleanup utility and now it has 600M which is plenty of room to install .NET and TurboTax. The .NET install always requires a reboot of the machine so it reboots and hangs on a corrupt pci.sys driver.

I don't know how the file went bad, but I couldn't even reboot in safe mode or recover from the Windows install disk.

The nice thing about Virtual PC's is that you can mount or add new drives. To fix this problem, I simply mount the drive to a different virtual instance and the bad virtual drive with pci.sys now shows up as D:. Copy the good pci.sys from C: to D: and the problem is solved.

All computers can crash but Virtual PC's are very easy to fix especially if you keep a backup on a DVD or make a copy just prior installing new drivers.

Bad Cables Often Crash Computers

My last Linux dedicated server managed to die an ungraceful death. I was fortunate to repair it three times by replacing the motherboard twice and the power supply once extending the lifespan of the server well into six years of 24/7 99.99% up time. But I had to say goodbye to the server and go a different route.

Usually after 3 years, parts become obsolete and the only viable solution is to upgrade your server. A quick trip to Fry's confirmed this.

I didn't want to build a new server from scratch so I created a virtual Linux server on a workstation to keep things up and running until the new hardware was purchased and shipped from Dell.

But the thing is, the virtual server actually ran well enough -- and perhaps better -- than the old server that died. After all, the server was running in its own logical processor that had more memory and about 2x faster that the six year old server.

So I didn't have to go out and buy new hardware at the moment.

The wait paid off since I was able to get a half-dozen used servers for $50 each without the drive. A local University was selling the two year-old computers -- minus the disk drive -- to get a newer model.

Now with 6 servers, I had enough spare parts to keep the server running for at least 10 years.

I gotta admit that virtual servers are pretty convenient, but the extra memory and disk space needed to run on my workstation really puts a limitation on some of the video editing and other application resource hogs. So I was somewhat glad to move the server over to a dedicated system.

I put in a new 500GB WD SATA drive, installed Linux Ubuntu 8.04 server got everything configured, rebooted and started getting SMART drive errors on boot.

I opened up the box, checked connections, same thing. I was ready to return the drive to Western Digital since all it had on it was the OS. I pull the drive out and boxed it up and closed up the server but I noticed that the SATA cable was kinked.

I replace the SATA cable and put the drive back in and all is working good!