Well, I think I got my WHS (Windows Home Server) box back up and running… for the most part.ADHERER

Does anyone out there have an unsuccessful Windows Home Server incarnation?  Seriously.  I read a good number of blogs by people who have their own WHS setup, so there’s plenty of company there.  But when there are issues, suddenly it is easy to feel like an orphan.

As near as I can tell, one of the drives in my WHS box (Acer EasyStor H340) went belly-up and just created havoc.  Some of the issues I experienced seem like they should be part of a widespread update.

First, as the drive was failing, WHS’ daily CHKDSK routine (which it saves the logs of) would grow larger and larger and larger.  Let me explain why this is a problem.

The Acer partitions the primary drive (1TB comes with the machine) with a small partition for the OS (comes up as C: on the server) and the rest is a seemingly-useless D: drive (useless because a recent update for WHS makes this primary data partition not keep any files… so this D: drive is constantly pretty well empty).  Well, of course, the CHKDSK logs are stored on the C: drive.  For 1000 points, can anyone tell me what happens when a Windows machine’s C drive fills up?  That’s right: chaos!  There’s suddenly no space for a swap file, so programs start dying with not enough memory to run in.  Start up takes forever.  Nothing seems to work right, and if it does, it takes a long time to get it to.  And to make matters worse, as soon as I would delete the offending log file, the server would create a new one by the next day.

I also learned that Windows’ indexing service was storing its large index on the C drive.  That was easy enough to change.  The documentation was actually helpful there.

But once I discovered a failing drive, I was stuck.  WHS wouldn’t allow me to remove the drive, since it couldn’t move the files off of it.  It also wouldn’t allow me to just pull the drive because that causes errors too, as the server discovers that files it thought it had are suddenly not accessible.  I was getting continual file corruption errors.

Eventually, what I had to do (and I don’t recommend this) was to manually delete everything on every drive in the box.  All the drives are mounted in the c:\fs folder, so it’s easy to find them.  I saved off what I could and just deleted the rest.  I lost backups of my networked computers as well as a ton of other backups.  Thank God they were just backups and not my primary/only source of information.

After totally destroying everything on the machine, I was able to get it back up and running… well, limping anyway.  I replaced the bad drive and now the server seems to be doing well.  The issue is that the server has many self-healing aspects to it.  But those procedures tend to rely on good hardware.  Once that’s no longer the case, the server’s self-healing procedures just made things worse and worse and worse.

I am very afraid that something might affect the primary drive of the server one day.  I really wouldn’t know what to do.  Acer didn’t give a decent manual with the machine, so it’s difficult to say what would happen.  There’s no method to back up the server, which is ironic.  The server backs up all the PCs on the network, but there’s no such backup for itself.  And it seems that I would need the bundled CD that came with it (wonder where the heck I put that) for anything.  But it’s not as thought the server itself has an optical drive.  Or a monitor port.  Or keyboard.  So for now I just keep my fingers crossed.