How to synchronize virtual machines with the cloud?

Question

Data synchronization

Good day. Actually the question is the following: how to set up synchronization of backup copies of virtual machines with some kind of cloud.
I work with three projects, each project requires its own stack, development environment, software, settings. I work on a fairly powerful laptop, but I would like to separate working projects by running them on separate OSes through a virtual machine. That is, to make a clean OS for each project with an installed stack and software for a specific project. The main OS on which the virtual machine will be launched is clean and not littered with the launch of many components of server software and development software.
I would be grateful if you share your own experience of setting up and synchronizing several projects using a virtual machine.
I repeat, the task is to isolate projects from each other and naturally from personal data and software on the computer, as well as a relatively simple way to recover data when reinstalling the main OS.

Answer 1

There will be no easy way. The root of the problem is redundancy and backup. The better solutions you need, the more difficult and expensive it is to implement them.
If you survive with backups after the shutdown of the virtual machine or less often, then it's simple, you can easily set up the launch of virtual machines from the local disk with a script, and upon completion of the work, the same script copies the data wherever you want.
I'm not saying what virtual machines are and what operating system, but the general principle is just that. Do you mount a disk with windows ntfs after completion of work, or do you have linux and all virtual machines work on a local network (even if everything happens within the same machine, a network is configured between virtual machines) with nfs boot ...
If you have linux, by the way, then all sorts of lxc / docker provide tools for organizing the workplace, with snapshots and so on.
If you have the opportunity, move the storage away from the working machine, but in the case of a laptop it will be slow wifi ... by the way, cascading data caching can turn out to be a beautiful solution ... i.e. you connect a block device over the network by adding (for example, linux bcache) a local disk to it as a cache, even in read-only caching mode this can give a huge increase in speed, coupled with the bonuses of independent storage (reserving which is easier to set up than on your laptop) will give you a great bonus.
-------------
An example of an organization, you have linux everywhere, on the server file systems with support for snapshots (for example, btrfs / zfs) all machines are loaded over the network, caching on local disks can be configured with something like cachefilesd / cachefs and so on, on the server your daemon creates according to your conditions snapshots (for copy on write file systems, this does not create overhead, at least do it every minute, if only there is enough space). Also try to transfer the database to the server, do not run them on a laptop.
Everything said by itself in the case of the organization of work through virtual machines can degenerate on one physical, i.e. a laptop is your server, and virtual machines are connected to it via a local network ... but all the risks of data loss arising from this do not go away.