Verge.io in Practice: A VMware Alternative from the Data Center

In our last blog post we described why Verge.io is a serious VMware alternative for many companies: fewer licenses, fewer systems, less complexity, and continued use of existing hardware. Compute, storage, networking, backup and disaster recovery on a single platform.

Sounds good. But how much of that really works as smoothly as it does on paper?

That is exactly what this article is about. Less marketing slide, more server room – in other words, what happens when someone actually stands in front of the hardware and says: "Right then. Off we go."

Getting started with Verge.io is pleasantly unspectacular. Instead of installing several software components, setting up various management servers and then marrying it all together, VergeOS is installed directly on the designated servers: boot the ISO, configure the network, connect the nodes – and individual servers become a shared cluster.

What is interesting here is not so much the installation process itself. What is interesting is what no longer has to be built afterwards: no additional storage platform that needs to be run and maintained alongside it, no separate network management for the basic functions, and no additional backup appliance as a system of its own with its own lifecycle.

That is the difference between "We installed a hypervisor" and "We built a complete private cloud platform".

The key question. Building a new platform is one thing – getting the 150 VMs that run on VMware today into it is quite another.

Verge.io comes with an integrated migration mechanism that takes over existing virtual machines. The data is not simply copied from A to B: changes that occur during the migration are transferred along with it.

Put simply, the VM keeps running while the migration is already under way in the background. Only the final switchover – the cutover – requires a short, plannable maintenance window. That reduces downtime considerably.

"Zero downtime" is still not a magic spell. Every environment is different, and dependencies, applications, databases, networking and backup all have to be taken into account. The decisive point is a different one: Migrating away from VMware does not have to be a multi-week major project. And that is exactly what suddenly makes VMware alternatives a great deal more interesting.

"Runs on x86 hardware" does not, of course, mean that every old server from the basement will do. In principle, Verge.io can be used on a wide range of off-the-shelf x86 systems: Dell, HPE, Supermicro and other platforms are just as much an option as individually assembled servers.

That opens up an interesting thought: Does the next server really have to be bought new?

Perhaps there are already machines in your own data center that would be perfectly adequate for the current stack. Perhaps the server is three or five years old. Perhaps it comes from an earlier virtualization environment. Perhaps it has long been written off, but technically it still has plenty of headroom.

Before a production installation, the specific hardware still needs to be put to the test. But the basic idea remains a smart one: the software does not care which brand is on the front of the server.

This is where the fun stops, at the latest. Backup is one of those topics nobody likes to talk about – until they urgently need it.

VergeOS integrates snapshots, replication and protection mechanisms against unwanted changes. Particularly relevant here are immutable snapshots: a stored recovery point can neither be modified nor deleted after the fact from within a compromised environment.

That is not a free pass along the lines of "We have snapshots, so nothing can happen to us". On the contrary – a well-thought-out backup and recovery concept remains indispensable. But the ability to protect and restore data and complete virtual environments within the same platform changes the way you handle disaster recovery considerably.

The dry statement "We have a backup" thus turns into the far more important question: How quickly are we back online after a really rotten day?

Sounds like a T-shirt slogan, but it describes rather well what changes in day-to-day operations. IT admins know the game: one console for virtualization, one for storage, one for networking, one for backup, one for monitoring, one for hardware – and somewhere in between a browser tab with the vendor's license portal.

Technically, all of that can be run perfectly well. How much time it costs to coordinate all these systems with each other is another question.

That does not mean there is no work left to do. But the number of systems, interfaces and dependencies goes down – and from our point of view that is one of the underrated advantages of this architecture. Fewer things to look after. Not spectacular, but exceedingly pleasant in everyday work.

On the first day, probably not much. A new platform has to be learned too: new terms, new menus, new ways of doing things. Once you are up to speed, one thing above all changes – the administrator has to mediate between systems far less often.

And anyone who wants to automate will find APIs as well as integrations with Terraform and Ansible. That also makes the platform interesting for everyone who no longer feels like doing the same 37 clicks by hand every Tuesday morning.

You do not install an operating system and suddenly all your IT problems are gone. Nor does the real added value lie in any single feature – not in virtualization, not in storage, not in networking and not in backup. It lies in the fact that these areas no longer have to be run as separate worlds.

That is why we find the approach exciting. Whether it holds up in your specific environment, however, can only be found out by putting it to the test.

Because for us, the most interesting form of technology is not what looks good on a slide. It is what still works for the administrator at 8:00 on a Monday morning.