C-Servers Announces Becoming Platform-Independent
cservers
Hosting Provider
Hello. At C-Servers, we always look for ways to improve our existing offerings and to provide the best possible market pricing for our customers at any point in time.
The recent development on market pricing in several fronts, with relevant increases accross purchasing several components and data-center space, along with some specific improvements we wanted to add to our offer, have led us to do a fundamental change on our most purchased product:
On the next days, C-Servers is departing from VirtFusion and becoming platform-independent.
We'll change our customers to a new platform, C-Servers EagleKey, with several technical advantages and an entirely different, simplified and fully integrated approach to our offering. We're bringing our most-sold product range in-house and adding to it extra possibilities, including Dedicated Servers (soon) and Resellers for our products (also very soon).
Over the past three weeks, over 140.000 lines of code were written and many tests and hardening were conducted, function-wise and security-wise, in order to guarantee our customers the best possible experience. Our chosen basis, .NET 10 LTS with Blazor, also allows us to do runtime changes with minimal to zero disruption at any point in time and to correct and react to any issues very quickly, which is an advantage. Our agent on the nodes, written in Go, allows our customers to have more performance available by not having to run PHP on every dedicated server.
We're also introducing several brand-new technologies for the benefit of our customers: this includes Cloud Hypervisor KVM with microVMs (Direct Linux Kernel) and several RAM improvements both from CHV and from us on this new launch. A Debian 13 VM under CHV, with the optimizations, has reduced from around 220MB to just 85MB without removing essential packages - providing you more room for what you wish to do.
Our NAT management also changes, from iptables to nftables, with improved response and technical conditions for our IPv4 NAT customers and a new button, "Flush NAT", automatically available for customers, in order to address a long-reported rare technical bug with connections on NAT. IPv6 provisioning will also start to be automatic on Zeta.9 Coventry and general dual-NIC servers as well.
The management interface at the WebStore will not change; it's only the Portal, and several underlying specifics, that change.
Most importantly: OS changes to CHV are opt-in, meaning that the templates are there should you wish to save on RAM on your system, but by default you'll still stay on QEMU/KVM; and the migration will incur zero server and zero IPv6 downtime, with a mere 15-second downtime for NAT IPv4 only due to the automated change from iptables to nftables, when it happens.
Lastly, this change allows us to publicly commit to this very important announcement: our low prices will be kept low. No increases for 2026 at C-Servers! We hope these news make you happy.
This is just a small summary on the upcoming improvements. Full details on the many, many technical specifics, reasoning and improvements are available here.
Let's C-Servers independently, shall we?
Have a great weekend!
Comments
Woohoo 🍻
Hi @cservers
I received a suspension notice stating: “Port scanning/worm propagation patterns: 172 different destinations per every 12h (average), 1/3rd unanswered (SYN_SENT).”
This VPS only runs Jellyfin, Navidrome, qbittorrent and rclone, and I’ve been using the same setup for about a year and a half without issues. I suspect the new C‑Servers EagleKey implementation may have flagged a false positive, possibly because I’m using a custom Alpine Linux installation. I’m happy to reinstall the VPS using the official EagleKey image if needed.
The performance of this VPS has been excellent, and I’ve had no complaints at all.
I’m posting the issue here because I’m currently unable to reach support through the web form.
Could you please review this for me? Thank you in advance.
Service - NanoVPS-II 768 DiskPlus (Zeta.6) FI (#1632) (vps-1632 - 192.168.240.207) - 14.63$
No - #1632
Hi, I'd also like some clarification around this. I usually do a custom install with Debian so I can implement LUKS. I don't know if what is mentioned about "direct linux booting" being a problem in this case. Can I still do custom installs on the new system? I'm concerned I'm locked out of my custom install method if I ever choose to reinstall
...congrats?
OpenVPN installer | WireGuard installer
(read it like a old Indian politician who dunno about what he's taking about) Big data, AI, machine learning
Hello @sicksaint - we don't provide support here for these matters. The support platform is up - you do need to go first to the Live Chat (available on the website or at support.c-servers.co.uk).
There's also a Knowledge Base article covering how to deal with the Live Chat correctly so that you can get to the ticketing section, if you're having difficulties with it.
But once we have the ticket we'll take a look at your case. The probable culprit will be qbittorrent over TCP, everything else seems quite reasonable.
Direct Linux Booting is essentially Cloud Hypervisor on the microVM variant (so CHV/KVM). You can do LUKS without any problem custom-installing via QEMU templates, yes.
Custom installs are available and will continue to be available, yes, but not from systems on "CHV Direct Kernel", due to the fact that microVM on CHV doesn't support loading ISOs on boot (lacks a UEFI for extra speed and RAM savings, can't load a drive on boot). You need to reinstall first with distributions that use QEMU.
We do not intend on ever restricting completely ISO utilization and we will continue providing templates on QEMU (and in fact extend them to other uses!). Any template on QEMU automatically activates the ISO capabilities.
You can go back and forth as well using QEMU and CHV templates - using CHV by default, and e.g. whenever you need to install something custom just reinstall with a QEMU template first, then you'll have the ISO capabilities available again on the panel (press F5). They're interchangeable.
Of course, if e.g. you intend to use QEMU for systems that have a CHV OS template, you'll lose the several I/O optimizations and RAM savings of CHV and our own design, but it's possible.
At the moment these are the templates on QEMU:
Everything else is Cloud Hypervisor.
Thanks so much for the clear explanation! In the post it wasn't clear to me if QEMU would just be gone if I ever reinstalled via template and couldn't get it back. Thankfully I'm not on a low ram plan so I'm not too concerned about the RAM saving, but that sounds super awesome for people who are! I'm always so surprised at how much mileage you can get out of a Debian install on a 512mb instance. All of this sounds very cool and thanks for continuing to innovate!
Thanks @cservers I have just created a support ticket.
Tiago from @cservers reactivated the VPS quickly and without any issues, showing excellent responsiveness and customer care. I would strongly recommend them as a solid choice for anyone looking for a cheap and quality VPS.
is qbittorrent use not allowed even say if seeding a linux ISO? what was the resolution
I was asked to use UDP over TCP for qbittorrent to avoid the flagging. In addition to this, I might need to reduce the number of peers in the qbittorrent settings panel.
This is terrifying.
love me some good old ai generated svgs lol
There's an issue with the management panel—none of the login methods are working right now.“Error.
An error occurred while processing your request.
Request ID: 00-b11b72c025266905722a6cf568f22173-fbeb8106ece8e05d-00
Development Mode
Swapping to Development environment will display more detailed information about the error that occurred.
The Development environment shouldn't be enabled for deployed applications. It can result in displaying sensitive information from exceptions to end users. For local debugging, enable the Development environment by setting the ASPNETCORE_ENVIRONMENT environment variable to Development and restarting the app.”
Sorted out since around a couple of hours ago.
We found these actually funny, so we let them stay
@cservers,
EagleKey seems to be working pretty well, although a bit slow for me. Otherwise, no issues!
I did want to make you aware that I received some emails about one of my services exceeding CPU usage. When I went to look, I had two services that were using more CPU than they should, and when I connected via console, they were both stuck booting with a kernel panic. I reinstalled the OS, which fixed the issue. Thought I'd let you know in case you're seeing this in other places.
Congrats
LumanexAI: Our in-house AI model for free 🚀
Glad things are smooth-sailing now, faster or slower.
As for the CPU - thank you for the heads up. That is a relevant question but, unfortunately, it can happen with any Linux machine, sometimes these do simply fail to boot... The advance warning is made (among other things) precisely for that, to allow for some time to correct any issues and for the user to intervene autonomously, and even 1st suspensions past that are not that harsh now (they're only 12 hours).
It's all much more progressive on CPU, disk and network usage management, and better - entirely automatic and seen 24/7. We've left entirely the previous system (where spaced one-off checkpoints with abuse were compensated with harsher suspensions or terminations). We decided to go softer there, you never know when someone needs to burst punctually because they're doing a backup or something similar, which is perfectly normal usage.
Those who will be noisy neighbours, will do it continuously or try to game the bursting system - there are checks and actions in place destined for both, steeming from our experience on those 2 years of operation. The others will get much better CPU for what they need. This is the main idea.
Just a notice, in the meantime: we did have a minor issue later today where 21 servers got inadvertedly suspended for 2/3 hours due to a minor math count fault on CPU percentage (hopefully yours wasn't caught on it). As soon as we seen that one we went immediately to sort out the fault. It's now fully fixed and the servers that weren't abusing (20) were immediately reactivated.
Thank you