We're investigating. Some active metrics might be stale.
The network provider has fixed the incident. It was congestion on one of the uplinks.
The write-path has been stabilised, meaning metrics are properly ingested again. We'll continue to stabilise the whole situation.
Investigations already started earlier this morning, while unfortunately at approx. 12:30 the situation worsened - impacting ingestion at that point.
Saturday 25th July 2026
Eindhoven (NL)Unreachability DNS servers
We are investigating.
The authoritative nameservers were under attack. They were able to absorb the attack, but our anti DDoS within the network started to block IPv6 traffic to protect the nameservers, even though they were able to absorb it.
We've tweaked the anti DDoS infrastructure to prevent this block from reoccurring. Only a select number of visitors were unable to do resolve hostnames.
