Earlier today, a brief service disruption occurred due to a misconfiguration on the upstream provider's end. Once downtime was detected, traffic was temporarily rerouted back to the previous router to restore connectivity. The upstream provider promptly resolved the issue, after which traffic was successfully routed as intended.

We are pleased to announce that traffic has now been migrated to a new router with significantly higher capacity than our current peak RX and TX utilization. This upgrade is expected to eliminate congestion occurring at the router level and provide improved headroom for continued growth.

As a next step, maintenance will be required on all cluster nodes at the location to implement upgrades aimed at resolving congestion on the node side. This maintenance will be announced in advance once the exact schedule is confirmed and is currently estimated to take approximately six hours.

During the maintenance window, all servers will be powered down and unavailable for the full duration. As this is planned maintenance, no compensation will be provided. The series of maintenance activities carried out over the past month is intended to fully resolve the connectivity issues previously experienced at this location and prevent a recurrence of the problems associated with the former upstream provider.

Further updates will be provided as preparation and scheduling progress.

A temporary switch to a new upstream provider in Finland has been completed. Prior to this change, the former upstream network experienced recurring packet loss, likely related to congestion within their infrastructure.

Traffic is currently operating through this temporary uplink replacement while stability is being closely monitored. The permanent switch/router hardware has already arrived at the data center and will be installed by technicians next week. Additional maintenance will then be required to migrate traffic from the current router to the new switch/router and finalize the transition.

Limited maintenance may also be performed to complete the upgrades associated with this network transition. Advance notice will be provided if such work is carried out.

Updates will be provided as progress continues.

The required switch has already been ordered, and we are currently awaiting for its delivery to the data center. Once the hardware arrives, our engineering team will begin deployment immediately, followed by configuration, testing, and preparation for the scheduled migration. We will continue providing progress updates as we move through each stage of the rollout.

Over the past two weeks, we have been closely monitoring multiple reports and network performance metrics indicating intermittent packet loss on routes transiting through our upstream provider in Finland. While overall connectivity has remained online, some customers have experienced noticeable issues such as increased latency during evening hours, packet loss on international routes, temporary instability for real-time applications (gaming, VoIP such as TeamSpeak and Discord, streaming, VPN, etc) and reduced network performance during peak congestion periods.

This behavior has been especially visible during peak traffic hours in the evening, when upstream congestion reaches its highest levels.

After a thorough investigation, we have confirmed that the packet loss is not caused by hardware faults inside our infrastructure, nor by issues within the data center itself. Instead, the root cause lies upstream - specifically, congestion occurring within the transit capacity provided by our current upstream provider in Finland.

During peak hours traffic volume increases significantly across the provider's backbone, leading to saturation of upstream links, queue drops and packet discarding, as well as increased jitter and unstable routing performance. Because this congestion happens outside of our direct control, the issue cannot be fully resolved through local solutions alone.

We understand that consistent connectivity is critical for all workloads, and we are treating this situation with the highest priority. To eliminate this issue once and for all, we are already implementing major network improvements.

We will be bringing up an additional uplink in Finland in the near two weeks. This will allow us to diversify transit paths, reduce dependency on a single provider, improve routing stability under load, and provide significantly more headroom during peak usage. As part of this transition, we will be discontinuing the use of the current upstream provider responsible for the congestion issues. Once the new uplink is fully deployed, traffic will be migrated away and we will no longer rely on that legacy upstream going forward.

To complete this migration we will require a scheduled maintenance window of at least 12 hours in order to safely transition our network to the new upstream provider and implement the necessary routing and infrastructure changes. This maintenance will be announced well in advance once we are fully ready, and we will provide all relevant details ahead of time to ensure customers can plan accordingly.

This upgrade is also part of our broader transition toward a Layer 3 network architecture. By migrating to an L3-based infrastructure, we will be able to remove bottlenecks caused by legacy transport constraints, enable better redundancy and failover between uplinks, and prevent congestion related packet loss from affecting customers in the future.

We recognize that this issue has impacted our customers' experience, and we sincerely apologize for the disruption. To compensate for the instability during this period, we will be providing service compensation to affected customers in a form of a service extension, proportional to the network disruptions.

We will continue providing transparent updates as deployment progresses. Once the new uplink is active and routing is migrated, network stability in Finland will be improve significantly, and all network-related issues will be gone.

Our goal is to ensure that issues like this do not occur again, and we are investing heavily into the network foundation needed to support reliable and high-performance connectivity for all customers.