Yesterday I made one of the mistakes I always warn my collegues about. I always tell them: “Make sure you know the situation before you change anything”. For some strange reason I made that mistake myself this time.
I just started a migration and got scared afterwards because the server gave all kinds of strange errors…
Together with Sander I did a couple of P2V’s for a customer. Normally I ask upfront if there are any problems known to the system and I reboot the system before doing the actual P2V. I also run through the eventlog to watch for any sign of serious problems.
Not this time. Why, I don’t know, but I didn’t.
After the P2V we got a ‘service not started’. Since I didn’t clean up the machine just yet it wasn’t that strange. It got strange after I removed all hardware related tools and drivers and still got the error after the migration. So after some searching and fixing things I finally asked the client: “Did this error occur before I did the P2V?” And guess what, yes, the error existed prior to the P2V. I should have checked it myself before I started.
This really annoyed me. I really like my servers (the ones I migrate that is) error free.
- Before you do a P2V make sure the physical servers are error free or that the errors can be explained
- Fix errors before you do the P2V
- Do the actual P2V
- Remove USB and serial ports after you did the P2V
- Remove all, and I mean ALL, hardware related drivers, utilities and programs
- Reboot and check eventlogs again
Hopefully I follow my own advice next time :(