Increased rate of SERVFAIL responses from DNS4EU node in Greece
Updates
Summary
Following a peering change in our server location in Athens, Greece on December 1st 2025 several domains became inaccessible due to non reachable authoritative servers.
In an effort to improve connectivity in South East Europe, we have decided to start peering with NetIX. This change was expected to improve the latency in the region and contribute to improving user experience.
Following the change, many authoritative servers became unreachable and DNS requests were responded with SERVFAIL.
We have investigated the root cause and identified that this occurred because of invalid routing. Requests toward authoritative servers were routed via newly established links, but responses weren’t delivered back properly.
We explicitly configured the communication with authoritative servers back to the default gateway which led to the resolution of the incident.
Timeline
29th Nov 2025 19:00 UTC: new peering with NetIX established. Monitoring started.
1st Dec 2025 09:00 UTC: Health checks succeeded leading us to believe that the node is healthy
1st Dec 2025 10:00 UTC: DNS Traffic started to flow towards the node
1st Dec 2025 11:00 UTC: Rate of SERVFAIL responses started increasing.
3rd Dec 2025 10:00 UTC: Issue discovered
3rd Dec 2025 11:45 UTC: Issue resolved
What we are doing to avoid this in the future:
Improved monitoring in order to discover increased rate of SERVFAIL responses
Fixed routing for queries towards authoritative servers
← Back