Minor incident
TCPShield: Brief partial disconnect in Frankfurt
Oct 25, 2026, 18:41 UTC
started
04:29 UTC
resolved October 26, 2026
9 h 47 min
duration
Affected: Frankfurt (Anycast)
Current status of TCPShield: Operational — live TCPShield status
Timeline
- Oct 25, 2026, 18:41 UTCidentifiedFrom 18:17 to 18:19 UTC, a single node in our Frankfurt fleet experienced an OOM (ran out of memory). Some traffic routed through Frankfurt saw a brief disruption of about 2 minutes. Service automatically failed over to other nodes in the process, and all other locations were unaffected. We have identified the root cause and deployed a corrective fix. No further impact is expected, and no action is needed from end users.
- Oct 25, 2026, 18:46 UTCmonitoringA fix has been implemented and we are monitoring the results.
- Oct 26, 2026, 04:28 UTCmonitoringThe affected node has been put back in rotation at approximately 03:30 UTC. A partial disconnect is expected in the process due to ECMP.
- Oct 26, 2026, 04:28 UTCmonitoringWe are continuing to monitor for any further issues.
- Oct 26, 2026, 04:29 UTCresolvedThis incident has been resolved.
Updates as published by TCPShield (original).
Other services with incidents at the same time
- Dartmouth College 19:26 UTC · Investigating Nework Issues at the Life Sciences Center
- Nexus Mods 18:03 UTC · Slow website / API responses
More TCPShield incidents
- September 6, 2026 Seoul optical issue
- July 6, 2026 Tokyo Hardware Failure
- June 30, 2026 Sydney outage
- March 30, 2026 Forwarding degradation in Amsterdam