We were able to recover cs1 at 03:26 and successfully completed the maintenance, including the pending cs2 reboot (03:29 - 03:41), followed by the necessary follow-up work till 06:00. A few minor tasks remain and will be completed tomorrow without customer impact, after which we will publish the announced maintenance summary.
Posted Sep 25, 2026 - 06:15 CEST
Update
DRT-FRA8 connectivity was cut at 01:11 as planned. One of the site's core switches (cs1) failed during the required firmware update and no longer provides serial console output. Customers connected to cs2 were restored at 01:58 via the new router.
Customers connected only to cs1 remain offline while we recover the switch or move their connections to cs2. To avoid risking the remaining operational core switch, we have deliberately postponed the cs2 reboot. Its update will follow only after cs1 has been recovered, so restored services may be interrupted once more.
Our priority is restoring directly connected DRT-FRA8 customers. Reintegrating the PoP into our network is secondary, as all other locations remain unaffected.
The maintenance window has been extended by 3 hours for now.
Posted Sep 25, 2026 - 02:11 CEST
In progress
Scheduled maintenance is currently in progress. We will provide updates as necessary.
Posted Sep 24, 2026 - 23:00 CEST
Scheduled
Start: 24.09.2026, 23:00 CEST End: 25.09.2026, 03:00 CEST
We will upgrade our DRT-FRA8 PoP by replacing the site's Juniper MX edge router with a new device. This will significantly increase our backbone and DDoS protection capacity at this PoP. Only customers connected locally at DRT-FRA8 are expected to experience downtime.
Who is affected: Directly affected (downtime): All BGP, GRE, IP Transit, and dedicated server customers connected locally at DRT-FRA8 (identifiable by the server's hostname). Indirectly affected (no downtime): All other customers. Temporary reroutings of some peering connections may result in slightly increased latency. Traffic and DDoS protection will be handled by our secondary PoP EQX-FR4. Our available backbone capacity will be reduced by roughly half but remains sufficient for the expected traffic levels. As usual, a residual risk remains when working on critical backbone infrastructure.
Fallback Plan: In the highly unlikely event of a critical failure at our secondary PoP before the new router is fully operational, we will immediately abort the migration and restore service via the existing router within 30 minutes. The same applies to any production issues with the new router.
Maintenance schedule & impact for DRT-FRA8: • Up to 30 minutes: We will disconnect the PoP from the dataforest network and validate correct rerouting behaviour. Local customers will remain online but without connectivity to other dataforest locations. To avoid blackholing, drain your services at this PoP at the start of the maintenance window if possible. • Up to 60 minutes: Full disruption for all local connections and servers at DRT-FRA8. In parallel, we will perform necessary software updates on the site's core switches to avoid a separate maintenance window. Depending on your connectivity, service may be restored and interrupted multiple times during this timeframe. • Up to 30 minutes: We will integrate the PoP back into the dataforest network. Once complete, connectivity will be fully restored and traffic can be routed back to this PoP.
The durations are estimates for each step and do not represent fixed start or end times within the maintenance window.
Customer action suggested (DRT-FRA8 only): • BGP, GRE & IP Transit Customers: Please actively reroute traffic away from your DRT-FRA8 sessions for the entire maintenance window to avoid disruptions or traffic blackholing. • High-Availability Workloads: We strongly recommend temporarily draining affected nodes from your global load balancing prior to the maintenance.
During the maintenance, our team will focus on the work itself rather than providing live status updates. We will report the actual duration of each step and any issues encountered once the maintenance is complete.
Posted Sep 21, 2026 - 00:28 CEST
This scheduled maintenance affected: dataforest Network (Edge Network (affects all locations), Dedicated Servers (EQX-FR4 & DRT-FRA8)).