Cloudflare status: hosting issues and outage reports

Problems in the last 24 hours

The graph below depicts the number of Cloudflare reports received over the last 24 hours by time of day. When the number of reports exceeds the baseline, represented by the red line, an outage is determined.

Most Reported Problems

The following are the most recent problems reported by Cloudflare users through our website.

• Domains (43%)
• Cloud Services (33%)
• Hosting (16%)
• E-mail (4%)
• Web Tools (4%)

Live Outage Map

The most recent Cloudflare outage reports came from the following cities:

City	Problem Type	Report Time
Istanbul	Domains	6 hours ago
Greater Noida	E-mail	3 days ago
Paris	Domains	4 days ago
Crisfield	Domains	5 days ago
Noida	Hosting	9 days ago
Augsburg	Domains	9 days ago

Full Outage Map

Community Discussion

Tips? Frustrations? Share them here. Useful comments include a description of the problem, city and postal code.

Beware of "support numbers" or "recovery" accounts that might be posted below. Make sure to report and downvote those comments. Avoid posting your personal information.

Cloudflare Issues Reports

Latest outage, problems and issue reports in social media:

• 
  whatnicktweets (@whatnicktweets) reported 6 minutes ago

  @boristane DO VPS => Never Cloudflare => Daily

• 
  Eli Edelkind (@eliedelkind) reported 6 minutes ago

  Fourth, cyber is a trust wedge into the enterprise. If a lab helps Cisco, Cloudflare, Microsoft, JPMorgan, Linux Foundation maintainers, or major software vendors reduce high-severity bugs, that buys credibility with the people who block or approve AI adoption. It is not just “sell a scanner”; it is “prove our agents can be allowed near your crown-jewel code and infrastructure.” Anthropic’s launch-partner list includes major cloud, software, finance, security, and infrastructure organizations; OpenAI lists major security organizations as Daybreak partners. Fifth, this helps with the coding-agent TAM, which is much larger than AppSec alone. Vulnerability discovery, patching, testing, repo comprehension, and dependency reasoning are all subskills of valuable software-engineering agents. The output may be marketed as “cyber,” but the learning loop improves the same capabilities needed for code migration, refactoring, testing, code review, and autonomous software maintenance. So, is there a better use of funds/time? For pure revenue maximization, probably yes: broad enterprise workflow automation, coding agents, customer support, analytics, office automation, and vertical business-process automation are larger markets. I do not think either lab is betting the company on AppSec. But as a strategic wedge, cyber is rational because it offers a rare combination: high buyer urgency, measurable outcomes, government relevance, safety justification, strong demos, and a proving ground for autonomous agents. The risk is that the marketing outruns the operational reality. CISOs should evaluate these tools as high-context AppSec/research accelerators, not as replacements for SAST, DAST, SCA, fuzzing, secure design review, human product-security judgment, or vulnerability-management governance. OpenAI itself still says SAST remains important and positions Codex Security around turning “this looks suspicious” into “this is real, here is how it fails, and here is a fix.” My bottom line: you are right to be skeptical of the hype; wrong to dismiss it as merely SAST. The commercial cybersecurity market is not the endgame. The endgame is proving, governing, and selling frontier agents that can safely operate inside the software-development and infrastructure stack.

• 
  Merlin (@TheWizardTower) reported 8 minutes ago

  @Ruff681368 If improper use isn't obvious, it's hiding a bug. The point of an abstraction is to make invalid states unrepresentable, and improper use obvious. This is what Rust/Haskell/Scala do by using Maybe/Optional instead of a universal null type. Even when CloudFlare had the outage because they called .unwrap() on a pointer, the meaning of that function is "You solemnly swear that this is a Some(x), because if it's a None, I'm blowing up your runtime." And, indeed, when that came up, the problem was *immediately* obvious. The go code in question here was code that looked reasonable at first glance, but wasn't. That's a problem!

• 
  Prashank Abhishek (@prashank25) reported 13 minutes ago

  @boristane Every day. @digitalocean @Cloudflare give me option to persist session for a week at least or make the 2nd factor remember the device because it's every login right now. Too annoying 🫩

• 
  Michael Kraus (@DataKraus) reported 13 minutes ago

  @Cloudflare You are the WORST! You ruin apps. You ruin websites! You make sure customers CAN'T get through. Viruses are more useful than your crappy Cloudflare software!!! I've stopped using several companies because of you. I hope companies wake up and ditch you and you cease to exist. You are the Ebola on the butt of the Internet. Please just go away!!!

• 
  father stretch my bandz (@MILKANDH3NNY) reported 18 minutes ago

  @boristane AWS - depends on what service Cloudflare - often but wrangler is great Vercel - often but mostly to look at usage (I refuse to pay them money) Fly - no need Railway - often; it’s pretty and usage is useful (just sorted a bug where they were charging for an orphaned container)

• 
  Vlaд Фāust (@vladfaust) reported 19 minutes ago

  @awesomekling @ladybirdbrowser Can you elaborate on the "help from Cloudflare"? Does it imply that if I am ever to write my own browser, I'd have to have some kind of approval from them so that the Captcha works? Thanks.

• 
  Flobert (@FlowHaa) reported 21 minutes ago

  @juancastillamar I buy my domains using Cloudflare and point them to my VPS IPs. No issues so far!

• 
  Solo (@projectsolo) reported 24 minutes ago

  At Consensus 2026, the hottest panel was titled "How to Prove You're Human in an AI World." Cloudflare and former U.S. officials on stage. 20,000 attendees. Proof of humanity is no longer a niche problem. It's the question the whole industry is now answering.

• 
  . (@r2fat3li) reported 27 minutes ago

  @Ahmeed2m @Bashmohandes From my understanding, caddy uses the cf api to renew certs, but the resolving is still done by cloudflare statically (does not change based on if I am connected to my home network or not) and only for public domains. I could buy a domain, but what about the static resolving?

• 
  Ayush Agarwal (@ayushagarwal) reported 27 minutes ago

  @dotmanish @dodopayments @Cloudflare yeah I was also shocked - stripe's website is the worst for agents.

• 
  CDSL – Central Depository Services India Ltd (@CdslIndia) reported 28 minutes ago

  @SHIVCHATTERJEE @AskZerodha Dear Valued Customer, thank you for writing to CDSL, please DM us the details of your public IP address and Cloudflare Ray ID, for us verify reason of blocked

• 
  John Monarch (@realjohnmonarch) reported 32 minutes ago

  @levelsio @Cloudflare I would love .bot too. But man the gTLD renewal fees are so bad. 4 letter “premium” .bot renewal fee is $600 a year.

• 
  Mark Harper (@jpr1chardson) reported 32 minutes ago

  Cloudflare just released its CLI tool for AI agents to easily manage infrastructure. The entire backend of the internet is being retooled to support agents. Now it just needs payments...

• 
  Nashon (@WEGNERJOASH) reported 36 minutes ago

  Worked on a project that can benefit those who want to lean customer based on their mpesa statement. You give them email (phone@domain). Use cloudflare workers to catch all emails. When the statement is received, the system request the client for the pin through sms....

• 
  Kevin Kern (@kevinkern) reported 36 minutes ago

  my little ios side gig has now an ai chat. the backend is built on cloudflare + better auth for google and apple sign in. alchemy is used for infra orchestration. codex is adding all my wishes and does automated ui tests in the simulator. i just have to yell at it when it goes off the rails. but i'm pretty happy with the results.

• 
  Kyle DH | pryvit.eth (@PryvitKyle) reported 38 minutes ago

  @CloakdDev IMO it’s early. As you point out most sites and data already make content free. x402 will primarily grow as Cloudflare turn their CAPTCHA systems into a toll booth and Google seems to be attempting to go down the ID route so they can place ads only for humans

• 
  Fredy Sandoval⚡️ (@devfredy) reported 39 minutes ago

  @asaio87 Will he secure your app behind a Cloudflare bot protection and set a Tailscale network for zero trust configuration, ensuring the database is unreachable for public internet?

• 
  Sattyam Jain (@Sattyamjjain) reported 43 minutes ago

  Five days ago, Cloudflare announced agents could buy domains. HN gave it 626 points and the reaction was "interesting experiment." Yesterday, AWS shipped a managed service for it. Concept → managed infrastructure in 5 days. Fastest we've ever seen this cycle complete.

• 
  Udit (@uditbatra1990) reported 45 minutes ago

  [quoting since can't reply to @levelsio] there's a middle ground that worked well for me. 1) add an extra "billing email" on whichever service allows it (eg. aws, cloudflare) 2) set up auto forwards to this billing email based on subject + sender for whichever service doesn't allow it you have to remember to do it *once* per service, and you end up w/ an email inbox full of just invoices

• 
  Shimecki (@scheemunai) reported 45 minutes ago

  @audiencon Or... you haven't implemented fallback to other providers. Otherwise what you're saying is - if your product dies when Cloudflare is down you're just a Cloudflare wrapper... or any other 3rd party that you're using in your stack. Including hosting, emails, or whatever else...

• 
  Lawrence Nwuzor (@Gozy_leonardo) reported 50 minutes ago

  I built my own MCP server in Python this week. Now Claude reads my Obsidian vault directly. 35 notes. 60+ folder indexes. Full markdown content. Searchable from any session. Stack: FastMCP. systemd. Cloudflare tunnel. 250 lines. The data never leaves my hardware unless I send it.

• 
  simulx4 (@simulx4) reported 53 minutes ago

  cloudflare rust edge workers are insane infra. holy crap.

• 
  Johnmark Obiefuna (@jayhemz) reported 55 minutes ago

  Proxying through Cloudflare solves almost all Wordpress security issues.

• 
  Bolt.new Help (@boltdotnewhelp) reported 55 minutes ago

  @thierrybezier @boltdotnew We recently experienced a service disruption caused by a Cloudflare issue, which has now been resolved. To ensure you're on the latest version, please perform a hard refresh of your project and log out, then back in. You should be good to go!

• 
  Build X Rajesh (@buildxrajesh) reported 58 minutes ago

  Google is slowly locking down parts of the free web while Cloudflare keeps blocking AI crawlers. Meanwhile half the AI agent startups are basically: “bro please let my bot read your website” 😭 The open web AI depended on for years is starting to disappear.

• 
  Somi AI (@somi_ai) reported 59 minutes ago

  @ayushagarwal @dodopayments @Cloudflare what's the routing signal, user-agent or accept header? UA spoofs trivially but accept: text/markdown would be the cleaner contract if agents start setting it

• 
  Bree Sharp | Local SEO Strategist (@bree_sharp) reported 60 minutes ago

  After moving MWOV from SiteGround to Cloudflare: Ahrefs went from returning 1 URL per crawl to crawling the site normally. The culprit was LiteSpeed's bot management — aggressive enough to rate-limit Ahrefs into giving up after the first URL. GSC showed 59 pages indexed fine. The problem wasn't the site. It was the layer between the site and the crawlers. Sometimes the issue isn't your content or your configuration. It's your infrastructure making decisions you don't know about.

• 
  Gelbooru (@gelbooru) reported 1 hour ago

  Gelbooru is down. Had to disable DNSSEC and wait for it to expire on all DNS resolvers. During this time, Cloudflare removed domain responses from their nameservers, which is causing this outage. They probably tried emailing us while the domain was unresolvable... Anyway, now that our DS key expired, we can migrate away with less downtime. Now it should be about 4-12 hours, instead of 24-48 hours.

• 
  Andrejco (@builtbyfaithh) reported 1 hour ago

  Today I finished setting up Cloudflare for a website. It was my first time doing the full process, so I was definitely a bit scared. Not because Cloudflare itself looked impossible. But because DNS is one of those t hings where one wrong record can break something important. • The website • Email • Webmail • cPanel • FTP So I was extra careful. I reviewed the DNS records Cloudflare automatically imported. Then I compared them with the original records from cPanel. Then I checked them again. And again. I made sure only the main website records were going through the Cloudflare proxy. • Mail • cPanel • FTP • webmail and other service-related records stayed DNS only so they would not break. I also found one small mismatch in the MX record and corrected it to match the original DNS setup. After that, I confirmed the A record IP matched the server IP in cPanel, collected the Cloudflare nameservers, changed them on the domain, and waited for propagation. At the start, it felt scary. By the end, it made a lot more sense.