Today I was working with one of my colleagues to upgrade an application we are running on some physical servers. We’re both working from home with a VPN connection into the data centre. Reboots were required. This is the bit I hate … a continuous ping times out for what feels like an eternity. Eventually that first response appears and the tightening of the chest relaxes 🙂
VM reboots are so quick because there is no hardware to POST. I could also take a copy of the VM to test the upgrade process before hitting production.
Aidan,
As you have VPN,why not us ILO/DRAC cards and KVM over IP. You can see al that is going on and have control when using ILO/DRAC. Beats waiting for that ping to return without even knowing if the system is booting or hanging somewhere in the POST sequence.
Take Care
Working for a hosting company, every piece of hardware has to be charged to a customer in one way or another. Customers are very cost sensitive and they won’t care if I am havnig a coronary or not. They care about saving a few Euros a month. Adding ILO means switch ports which means money so we don’t do it on rack servers. We have it on the Hyper-V cluster blades but that’s it. We trim every penny from the luxuries and focus on doing the essentials as good as possible.