Automatic mirroring/backup using snapshots on the runtime can minimize the number of fail-recovery procedures and down time overall. Running Virtualboxes with individual stacks clarifies use cases and minimizes risk while users test different partial solution elements.
Virtualboxes can also run in different physical machines so that the relative demand of processing power can be met while maintaining optimal usage of each physical machine. This ensure that the risk of overloads is minimized during peak demand times.
Improvements to My Organization:
It's helped us utilize spare capacity on servers and test new platforms and frameworks. It has also improved the staging of our apps before implementations.
Room for Improvement:
Managing virtual machine stacks on an individual and stacked basis needs work. It also lacks the ability to perform auto-recovery switchover.
Moreover, they should define Virtualbox stacks as groups of virtual machines working together as solutions. It would also help if they could define auto-backup/mirroring and failover configurations.
Use of Solution:
I am still evaluating the methodologies with the current free versions.
We've had no issues with the deployment.
We've had no issues with the stability.
There have been no issues scaling it for our needs.
I set it up myself and tried to configure it to the extent of my knowledge.