Hi Sectigo folks,
Yesterday morning (US time), we started encountering a large spike in 404 errors coming from sabre2025h1. This has resulted in not just significantly dropped log availability, but many certificates violating MMD. I've attached two arbitrary examples, but we've collected several hundred, both from certificates generated by our monitoring infrastructure and in-the-wild certificates.
Please stop this log from accepting additional submissions ASAP until such time as you have confidence that the log will be able to fully recover and include all outstanding and subsequently submitted certificates in a timely manner.
Please acknowledge your investigation on ct-policy@ as soon as possible.
As you learn more, please provide regular updates, info on whether the log will be able to fully recover, etc., to ct-policy@
Thank you,
Joe, on behalf of the Chrome CT team
Based on our monitoring, and a notification from the Chrome CT team, Sabre2025h1's MMD was exceeded between 18:17 UTC on the 26th of March and 22:29 UTC on the 27th of March.
In the last few days, we've seen a large decrease in usage of sabre2025h1, with a similar increase in usage on sabre2025h2 as 90 day certificates now fall within that log shard's window. So far, sabre2025h2 is coping.
We keep monitoring this, but as we mentioned previously, we view the best path forward for the long term resolution on these issues to be the usage of our Postgres-backed CT Logs.
Meanwhile, we may further tweak rate limits, especially for submission of requests. While we're not particularly fond of having to restrict submissions, between risking an MMD violation or stricter rate limits, the latter seems the lesser of two evils.
Regards,
Martijn Katerbarg
Sectigo