We will be carrying out urgent security maintenance across 1/3 of our VPS nodes to address a critical security matter and keep our infrastructure protected.
A reboot of all VPS instances hosted on the impacted nodes will take place during this maintenance. As a result, customers should expect a brief service interruption of up to 15 minutes per VPS while their server reboots. No further impact is expected beyond this short reboot.
🕐 The maintenance is scheduled for July 7th, 2026, from 11:15 PM UTC and is expected to be completed on July 8th, 2026 by 3:00 AM UTC.
No action is required from clients. All virtual servers will automatically come back online once they have been rebooted.
This maintenance affects VPS nodes only. All of our other services remain fully operational and are not impacted in any way.
Impacted cluster nodes:
🇩🇪 Frankfurt (fra01)
- fra-9950x-pallida
- fra-9950x-rooki
- fra-9950x-otocyon
- fra-9950x-zerda
- fra-9950x-velox
- fra-9950x-ruepellii
- fra-9950x-vulpes
- fra-9950x-urocyon
- fra-9950x-corsac
- fra-9950x-cana
- fra-9950x-ferrilata
- fra-9950x-odessana
- fra-9950x-lagopus
- fra-9950x-bengalensis
- fra-9950X-macrotis
- fra-9950x-chama
- fra-9375f-budget-2
- fra-6140-budget-7
- fra-3700x-storage-1
- fra-9575f-zerda
- FRA Game 1
🇳🇱 Amsterdam (ams01)
- ams-9950X-velox
- ams-9575f-macrotis
- ams-9950x-fulva
🇫🇮 Helsinki (hel01)
- hel-7950x3d-rueppellii
- hel-7950x3d-macrotis
- hel-9454p-otocyon
If you experience any issues following the maintenance window, please don't hesitate to reach out to our support team for assistance. You may contact us through tickets in the Client Area, Telegram support bot or over e-mail.
Thank you for your understanding as we work to keep our infrastructure and customer servers secure.
--
04:00 AM UTC
The urgent security maintenance is now fully completed. All cluster nodes have been patched, rebooted, and are operating normally.
Rather than taking everything down at once, we worked through each cluster node one by one - this kept simultaneous downtime across affected nodes to a minimum and let us apply the necessary changes precisely on each host. Because of the nature of the fix, some nodes required structural changes and took longer to process than others, so thanks for bearing with us while that was done carefully across the fleet.
This addressed a recently disclosed Linux kernel vulnerability, Januscape (CVE-2026-53359). In simple terms: it's a flaw in KVM, the technology that keeps each virtual server isolated from the physical host and from other customers on the same machine. Under specific conditions, it could let a malicious VM break out of that isolation and compromise the host, if nested virtualization was enabled on the host. It had gone unnoticed for ~16 years (!) and affects both Intel and AMD systems.
We have no indication it was ever exploited on our infrastructure — this was a proactive patch applied as soon as fixes landed, which required a kernel update and reboot on each node.
No action is required from clients. Everything is back online.
If you run into any issues, reach out to our support team and we'll be glad to help.
Arelion has confirmed that the problem was on their side, caused by congestion issues within their network in Finland.
Our Finland network is now operating normally, and packet losses are no longer present. We will continue to monitor the situation to ensure stability remains consistent.
Fun fact: this was most likely linked to the UEFA World Cup, with a live game between Belgium and Egypt taking place right now.
We thank everyone for their patience and understanding throughout this incident.
The issue has been traced to Arelion's side, with packet losses visible directly on their routers. This confirms the problem lies within their network and not our own configuration. This issue also impacts many other providers who rely on Arelion. Our upstream provider has escalated the matter, and we are currently awaiting resolution on Arelion's end.
As the fix is dependent on Arelion's team, we do not have a definitive timeframe at this stage. We are staying in close contact with our upstream provider and will provide updates as soon as the situation progresses.
We apologize for the disruption and appreciate your understanding. Further updates will follow as soon as we have more information.
As we promised, we are providing a full post-mortem of what happened in order.
What happened
It all started three months ago, when our Zet.Net uplink came under a large-scale DDoS attack, heavily saturating our uplinks and putting our Netherlands location as a whole at risk. This was also heavily compounded by ongoing issues with Zet.Net themselves - their services had been severely declining, with often congestion issues during peak hours, as well as their NOC team not being very responsive. They were also unable to provide a solid solution for our network at that time. Given the attack combined with the overall deterioration in quality, we made the decision to cancel our contract and reroute all Netherlands traffic through our Germany location. While this added a few milliseconds of latency, it allowed us to protect the servers while we searched for a better long-term solution.
During this period, we were actively on the lookout for a new upstream provider - one that could properly accommodate our Netherlands location and allow us to live up to the promises we've made to our customers.
Once we found the right provider, we began the migration. Due to a combination of issues that arose during the process - including mistakes made on both the upstream provider's end and our own - the transition did not go as smoothly as intended. As a result, some of our traffic was unable to reach our servers properly, causing network drops and, in some cases, downtime for affected customers.
Resolution
We're pleased to confirm that the maintenance has already been completed, and our Netherlands network is now running on the new upstream provider. The new upstream provider features a large number of private peerings with key providers across Central and Eastern Europe, delivering even lower ping times for our customers. This upgrade also lays the groundwork for future development of our network infrastructure, including the planned deployment of a core network in one of our key data centers.
Compensation
We understand these issues were frustrating and caused real disruption for many of you. To make things right, we will be providing three additional days of service compensation to all impacted customers.
Conclusion
Looking ahead, this upgrade will allow us to further develop our tariff plans and expand our range of offerings, including custom solutions tailored to your needs.
We thank everyone for their patience throughout this incident.
We've managed to bring these plans back at our usual fair pricing, without passing on the inflated component costs to our customers. Stock is limited, however, so we'd encourage anyone who's been waiting to order sooner rather than later, as availability may fluctuate.
Please note that our Ryzen 9 9950X lineup in the Netherlands remains out of stock for now. We're continuing to monitor the situation closely and will provide a further update as soon as those plans can be restocked under the same pricing.
Thank you for your continued patience and support. Further updates will follow as the situation develops.
This shortage is driven by a worldwide surge in demand for RAM and storage, which started in the fall of last year. This surge is largely fueled by the rapid expansion of the AI industry - as major players acquire memory and drive capacity at scale, availability across the market has tightened significantly, and component pricing has risen quite a lot as a result.
We want to be fully transparent about why we have chosen not to simply restock these plans at this time: at current market prices, bringing our plans back into stock would effectively force us to raise prices by at least 40% for new customers. This is something we are not willing to do. We believe in keeping our pricing fair and consistent for everyone, and we do not want new customers paying a premium driven by temporary market conditions.
At this time, we are unable to provide an estimate of when our plans will be back in stock, as this very much depends on the hardware market and pricing stabilizing. We are monitoring the situation closely and will restock as soon as we are able to do so without compromising on our fair prices.
We appreciate your understanding and patience during this period. Further updates will be provided as the situation develops.