Story goes…I was trying to consolidate my updates into three easy parts.
All Linux instances to be updated via Landscape
All Windows Server instances to be updated by WSUS
Both Windows 10 instances via WSUS.
It kinda worked – until it slowed TV-Server down to a crawl, My Windows 10 instance couldn’t find Physical Server to update from and I had mucked up Web-Server, TV-Server, Kodi-Server and Windows 10 instance in all this. A well as WSUS being RAM hungry! (if you are gonna mess up, do it well!)
Damn… all that restoring! err….no…
You see WSUS is ‘only’ a role so delete that of Physical Server and that’s done. No more RAM hungry roles on my server!
That leaves ‘only’ the four VMS that I mucked up. I take regular exports (usually when I’ve done something major and it’s worked!) of the VM’s so it meant I could go back to what I had before. This was as simple as copying a file to the VM Drive folder. (Virtual Hard Disks are just a file!) Do that four times and BANG! four happy VM’s.
Now – this is a blog, so it meant a couple of pages were missing. Now my backup plan is this for data:
Kodi Server backs up at 3am any new media files to an external drive
Email Server backs up at 4am all email domains and addresses to a shared external drive
Web Server makes a backup at 5am locally and then at 6am it throws a copy onto Email Server’s external drive.
So, a simple copy of the archive of those websites (databases and all) and run the restore from the command line – I got everything back!
This meant after my cock up, within 1/2 hour I had everything back to the way it was!
This is why we back up if an update goes wrong, or a feature you put in doesn’t work, or malware gets into your system, then you can hopefully get back to where you were before things go wrong, (or you can’t get something to work!)
(This was a post from my earlier blog – which I liked the post – so resubmitted it for here! – and for reasons notwithstanding – been moved again!)