In this article
Let's Encrypt loses outbound IPv6 after datacenter maintenance
Let's Encrypt took one datacenter offline on 3 November 2026 for scheduled maintenance, including an upgrade of its internet uplinks to higher-speed fiber. Afterwards, outbound IPv6 from that site, called HA Datacenter 1, broke, and the certificate authority says about 50% of domain validations over IPv6 failed.
IPv6-only subscribers saw half their validations fail
The failures hit subscribers who validate only over IPv6, with no IPv4 fallback, from 4 November 22:30 UTC to 6 November 16:30 UTC, according to Let's Encrypt. Dual-stack subscribers were largely unaffected because validation fell back to IPv4. Let's Encrypt says most subscribers also have IPv4, and fallback code in its validation software, Boulder, limited the damage.
Staging was affected, production API moved to Datacenter 2
Staging was affected for the whole incident. Production API traffic was shifted to HA Datacenter 2, which Let's Encrypt says worked as intended. If your ACME client runs on a host that only uses the newer internet protocol, check its logs for failed validations in the affected window and renew again.
Let's Encrypt apologizes and plans network monitoring
Let's Encrypt says it did not notice at first. IPv6-only subscribers are rare, fallback code masked the failures, and some black-box monitoring failed without pointing to IPv6. Its main takeaway is the need for better monitoring and alerting on lower-level network connectivity and on the IPv6-to-IPv4 fallback rate. It apologized for the downtime and thanked subscribers for engaging on the community forum.
From uplink upgrade to resolved write-up
- 3 November: the datacenter went offline for maintenance and the uplink upgrade.
- 4 November, 22:30 UTC: the failure window for IPv6 validations began.
- 6 November, 16:30 UTC: the failure window ended, according to the post-incident summary.
- 6 November, 16:52 UTC: Let's Encrypt identified the IPv6 problem, contacted the upstream provider and moved production API traffic to Datacenter 2.
- 6 November, 17:17 UTC: it was waiting on a fix from the upstream ISP.
- 6 November, 21:22 UTC: a fix was executed with the provider and traffic began shifting back to Datacenter 1.
- 16 November, 18:18 UTC: the incident was marked resolved with a write-up. Let's Encrypt says several community forum reports of failing IPv6 validation led to detection, and it moved traffic out before engaging the ISP.
An upstream ISP misconfiguration broke IPv6
Let's Encrypt names an upstream ISP misconfiguration in one of its datacenters as the root cause. After the fiber upgrade, the new network first worked over IPv6 but not IPv4. The ISP's fix for IPv4 then broke IPv6.
What Let's Encrypt has not published
Let's Encrypt has not said how many subscribers or certificate issuances were affected, and it has not named the ISP.
Sources
Image: Photo by Markus Stickling on Unsplash
This story was produced with AI assistance from the sources listed above and fact-checked against them before publication. Read our editorial methodology and the corrections log. Spotted an error? Email [email protected].

Discussion