Update: we're still seeing 2.1x as many sites being down as our normal baseline, not every site has been brought back online yet 👀
Where we could safely detect it, we sent out e-mail alerts to team owners if we detected their monitors were still running old/unpatched PHP versions 💪
Stay safe out there folks - patch asap! 🙏
These last 24hrs have been interesting in PHP land, to say the least! Here's everything we saw, from a monitoring perspective.
At its peak, 18.8% of all sites we monitor were down 😱
For the servers that answered with an HTTP/503, the majority was behind Cloudflare.
In hindsight, that might have been unnecessary. Cloudflare was very fast to deploy mitigation via their WAF rules for all customers, but at the time, we couldn't have known for sure. Safety first!
You can read more about these in our blog, with some more commentary & interactive charts to see more data visualisation on these last 24hrs ⬇️
We'll update the post in a few hours again to see how the tail-end evolves and if those sites recover 🤞
ohdear.app/news-and-updates/…
Out of caution, we've taken our website & dashboard offline, until the official PHP patches come out.
All your monitors continue to run in the background & you'll still be alerted.
Stay safe out there, fellow PHP sites 🖖
Oh Dear retweeted
Out this week:
Simpler first monitor setup, webhooks that fail for 48 hours now pause themselves, scoped API tokens locked out of team-wide endpoints, stricter email verification, and a bunch of fixes, including missed scheduled task pings. 🔥
ohdear.app/changelog/2026-10…
Out this week:
Simpler first monitor setup, webhooks that fail for 48 hours now pause themselves, scoped API tokens locked out of team-wide endpoints, stricter email verification, and a bunch of fixes, including missed scheduled task pings. 🔥
ohdear.app/changelog/2026-10…
Oh Dear retweeted
Small UX improvement to @ohdearapp: our MCP path at /mcp now shows a quick placeholder with instructions, if you come via a browser ohdear.app/mcp
For most projects, that's just a broken link (unless used via a coding agent), feels like an easy win 👀
New week, new changelog.
Configurable check frequency for Application Health, stronger SSRF protection on uptime and AI checks, task details in late scheduled task webhooks, and a bunch of server errors turned into proper validation messages. 🔥
ohdear.app/changelog/2026-10…
Latest changelog is out.
Up to 90 days of uptime history on status pages, custom status page icon colors, uptime check intervals up to 24 hours, a fix for abandoned status page CNAMEs, and a bunch of fixes. 🔥
ohdear.app/changelog/2026-10…
Here's what landed this week:
Scheduled tasks can now be added to server monitors, Lighthouse checks now run from a consistent server per continent so results stay comparable, and a bunch of API and port check fixes. 🔥
ohdear.app/changelog/2026-09…
Quick recap of the week:
Refreshed designs across the feature pages, a clearer GDPR FAQ on procurement and data processing agreements, and a fix for status page logo uploads that always failed validation. 🔥
ohdear.app/changelog/2026-09…
Oh Dear retweeted
Same for @OhDearApp - always disliked per-user pricing. Heck, we’re even leaving money on the table by offering SSO to our most basic customers - no enterprise or security upsell !
This week at Oh Dear:
Our MCP server gained a pile of new tools for downtime, maintenance windows and status pages, Sitemap alerts got a consecutive-failure threshold, and Lighthouse checks now actually run daily. 🔥
ohdear.app/changelog/2026-08…














