I have been using OVH for around 13 years, mainly for dedicated servers.
My usual setup is simple: I rent the physical server, install Proxmox, and build the actual infrastructure using virtual machines and LXC containers. Docker may also be part of the environment when it makes sense.
Over time, I have used standard OVH dedicated servers as well as SoYouStart and Kimsufi.
My opinion is straightforward: for the combination of price, hardware and stability, I have personally found OVH extremely difficult to beat.
My typical setup: dedicated server, Proxmox, VMs and LXC
I rarely rent a dedicated server just to run a single application directly on the host.
In most cases, I install Proxmox and use the physical machine as the foundation for multiple isolated systems. Depending on the project, this may include Linux virtual machines, LXC containers, web servers, databases, mail systems, firewalls or Docker environments.
This gives me control over the physical hardware while allowing me to separate workloads and allocate resources according to actual requirements.
For this kind of infrastructure, OVH has worked extremely well for me.
OVH, SoYouStart and Kimsufi
I have used all three extensively and choose between them according to the project.
Not every workload needs the newest CPU or expensive enterprise hardware. Sometimes a lower-cost dedicated server is perfectly adequate for running Proxmox, several VMs, containers or a lab environment.
This is one of the strengths of the OVH ecosystem for me: different options for different budgets and workloads.
The price-to-hardware ratio has often been excellent.
Stability is where OVH really stands out
Cheap hardware means very little if the infrastructure constantly creates problems.
The only major outage I clearly remember was the exceptional event when an OVH data center caught fire years ago.
Apart from that, my experience has been remarkably stable. I cannot remember recurring infrastructure problems or a pattern of downtime that made me reconsider using the provider.
The servers simply run.
That is exactly what I want from infrastructure.
I have never needed technical support
I cannot honestly review OVH technical support because I have never needed to contact them.
I install and manage Proxmox, networking, virtual machines, containers and the services running inside them myself. I have simply never encountered a provider-side technical problem that required me to open a support ticket.
So I do not know how quickly support responds or how good they are at solving complex issues.
I would rather say that clearly than invent an opinion about a service I have never personally used.
What I do not like: IP reputation
The weakest part of my experience with OVH is IP reputation.
In my experience, newly assigned OVH IP addresses are very often already listed on one or more blacklists.
For many workloads, this may have little practical impact. For mail infrastructure, it can be a serious problem.
You can configure a server correctly, secure it properly and follow good operational practices, but still inherit an IP address with a bad history.
Because of this, I never assume that a newly assigned OVH IP has a clean reputation. I check it.
This problem is not unique to OVH, but OVH IP space does not have a particularly good reputation in this area. In my opinion, the company could probably be less permissive with abusive customers and problematic activity.
Legitimate users should not have to inherit the consequences of someone else’s abuse.
Would I recommend OVH?
Yes. Absolutely.
For dedicated servers running Proxmox, virtual machines and LXC containers, OVH remains one of the providers I trust most.
I also continue to consider SoYouStart and Kimsufi when lower-cost dedicated hardware makes sense for the workload.
The biggest strength is the combination of price, hardware and stability. The main weakness, in my experience, is IP reputation.
That is the honest balance: excellent dedicated infrastructure, remarkable stability, strong value for money, and IP addresses that I always check carefully before putting reputation-sensitive services into production.