2026-10-16 09:37 UTCIncidents Detected

Cloudflare Access Issue Ongoing

Cloudflare is working on a problem with Cloudflare Access, its Zero Trust login gateway. Elsewhere, the major cloud and platform providers we checked report normal operations. We found no internet-wide outage, major BGP event or DNS failure.

**Active: Cloudflare Access availability issues**

Starting at 09:20 UTC on October 16, Cloudflare has been investigating problems some users hit after logging in through Cloudflare Access. One minute later it marked the cause as identified and said a fix was being implemented. The incident is rated minor. It affects organisations that put internal apps behind Access, not Cloudflare-proxied websites in general. As of about 09:40 UTC it had not been marked resolved. Source: [Cloudflare Status](cloudflarestatus.com)

**Resolved in the past 72 hours (Cloudflare)**

**Eastern North America, October 15:** Cloudflare saw network performance problems at its Ashburn (IAD) and Newark (EWR) facilities, with elevated latency and errors from about 20:48 to 21:38 UTC. The incident was rated major but was brief.

**Increased HTTP 502 errors, October 15:** Cloudflare opened an incident at 10:47 UTC and marked it resolved at 20:39 UTC.

**pages.dev in Norway, October 15:** pages.dev subdomains were unreachable from Norway from about 16:42 to 21:26 UTC. Pages sites on custom domains were not affected.

**Workers Cron Triggers, October 16:** Some customers' cron triggers did not fire between 00:01 and 01:06 UTC.

**Seoul (ICN) and Bangkok (BKK), October 16:** Regional congestion raised latency in Seoul (01:13 to 02:11 UTC). Bangkok saw slower response times and some 5xx errors (06:48 to 07:30 UTC).

**Other notes**

**Twilio** is still tracking several minor, carrier-specific SMS, MMS and voice delays, including Brazil, Thailand, China and Australia routes. These are long-running and narrow in scope.

**OpenAI** shows minor delays to Costs data in its Compliance API. ChatGPT and the main API are not affected.

**Regional ISPs:** Cogeco (Ontario) had a roughly two-hour landline outage on October 12, now resolved.

**All clear:** AWS, Google Cloud, Azure, Azure DevOps (routine maintenance only), GitHub, Akamai, Anthropic/Claude, Discord, Atlassian, Slack, Zoom, Vercel, Heroku, DigitalOcean, Cursor, Groq, ElevenLabs, Cohere and Stability AI all reported normal operations. We could not read the status pages for Stripe and Meta (Facebook, Instagram, WhatsApp) automatically. Searches found no outage reports for either today, but we have not confirmed their status directly.

The internet is up.

← Home View all history

Cloudflare Access Issue Ongoing

Last 30 Days