In one of our labs, we have (had?) a moderately powerful machine. It was a Dell dual core server with 16GB of RAM and 5 total hard drives. We used it to host virtual machines that ran SharePoint and DAV servers, which we in turn use to test OneNote.

Coming in this morning one of the testers noticed the machine was down. Log in remotely - no good. He went to check on the machine and noticed that it was dead. The first clue as to what happened was that the fan would not come on when he applied power and an odd clicking noise from one of the hard drives. And anyone that has ever heard that clicking noise from a drive knows the outcome is usually grim.

Closer investigation showed a few drops of oil on the fan blades. The cooling fan is nothing special, so those drops of oil seem very out of place. None of the capacitors in it seemed to be oil filled.

Upon turning it on, the POST routine runs for a few seconds, then it freezes.

Fortunately, it is still under warranty so we should be able to get whatever ails it fixed.

Here's the analysis for what we think happened.

  1. At some point over the weekend, the fan stopped. I don't know if this is due to the oil, or the oil droplets are a factor as to a different problem that also took out the fan.
  2. The machine overheated.
  3. We wanted to run a RAM test to see if there was a problem with the SIMMS, but the BIOS does not respond long enough to start the test.
  4. At this point, there is little else we can do. Either the motherboard or the CPU is dead, so we will just wait for the tech to replace it.
  5. No idea why the hard drive was clicking - overheating or getting oil on it are two culprits.

Each of the other hard drives held a different VM, and they all still work. I don't think we will lose any data here - just time waiting for parts. The clicking drive held the host OS, which fortunately is not that critical in this scenario.

What a silly way to start a work week.

Questions, comments, concerns and criticisms always welcome,

John