How to secure data on Hyper-V?

Question

Hyper-V

Colleagues,
the question is,
there are 2 servers on which Win Server Hyper-V is installed,
I want to make it safer, and so that if one Server dies abruptly, the data will not disappear anywhere, and it’s also desirable that virtual machines don’t stop,
now Virualks work on one server, and on the second everything is replicated.
Are there safer configurations?
As far as I understand, HA Cluster on two nodes (Without external storage) cannot be raised?

Answer 1

The data is not stored in the same place where the server works, regardless of the type of virtualization system, the containers themselves are kept on a third-party NAS, otherwise all sorts of reliability features such as High Availability or cooler will not work - Fault Tolerance
ps for your task - if you had a host linux, you could put DRBD nodes on each server - distributed storage, and already keep your images on it (virtualization systems do not directly support DRBD, but linux can work with them natively, while the driver is being sawn for windows ), in any case, this bad option.

Answer 2

You have hyper-V. Use replica. This is both a backup and a copy preparing for battle.

Answer 3

You are making a big mistake if you think that "suddenly the server will die" ...
First, you must proceed from the fact that he will definitely die, and for this he will choose the most inopportune moment.
Secondly, besides hardware or software failure, there are many more beautiful scenarios of the apocalypse, and the probability of them - you multiply by zero in vain.
It can, for example, turn off one of the RAID disks and throw out a portion of data mixed with garbage, and this will damage the file structure and, in general, all information.
It may very well happen that, for example, a ransomware attack will happen, and a topic that has been popular recently is that not a virus will work, but a living person. And this man, #c%@, will do everything correctly. And you will lose: hypervisors, and virtual machines, and domain controllers, and shadow copies, and replicas, and data on remote servers, and cloud storage will bypass and crash. And even if they don’t crash - while you are re-creating the entire infrastructure, pulling data from the clouds in an hour-by-teaspoon - these are weeks of downtime. It will be easier to pay the ransom.
Set yourself a task - what to do with such an introductory? And decide.
There are solutions, simple and inexpensive. In any case, their price (in case of management greed) is lower - than from the consequences for business without them ...
Imagine a Monday morning: the servers are fully encrypted, the customers are standing in line at the warehouse, some of them have already paid for the goods, the supplier has brought a truck - it’s also worth it ... But you don’t know what you have in your warehouse, at what price, what paid, what not, who you owe, who you don't.
In our time, it is better to be a trained pessimist than a naive optimist.