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 (36%)
• Cloud Services (34%)
• Hosting (25%)
• Web Tools (4%)
• E-mail (2%)

Live Outage Map

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

City	Problem Type	Report Time
Ashburn	Domains	24 hours ago
Rosario	Domains	5 days ago
Merlo	Domains	7 days ago
Frankfurt am Main	Hosting	7 days ago
Birmingham	Hosting	11 days ago
Dayton	Domains	11 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:

• 
  Dani Pralea (@DanutPralea) reported 2 minutes ago

  Your Cloudflare default settings are probably fine. But if you ever enable "Bot Fight Mode" and notice your signups drop, check your event logs - it flags VPN traffic aggressively and you can't easily see it in analytics. Happened to me this week while debugging a different issue.

• 
  Lee Penkman (@LeeLeepenkman) reported 2 minutes ago

  @garrytan @jack @tobi same with cloudflare is gmi because they are vibecoding vinext and all this... have lots of servers too... its interesting adobe and autodesk i think someone needs to talk to that because they are doing art/3d AI models that are quite unique. i can see why adobe going down though because of crime against their customers ofc. but im interested to see if autodesk makes a comeback too. Optimizing AI for construction is going to be massive. same with entertainment... Autodesk needs to vibecode an unreal engine and a marketplace for that .... if they havnt already... and aquire a company like fal and go into AI datacentres :D sometimes i wish i could just take over some of these companies and try quickly turn the ship lol.

• 
  if𝑒 (@ifeanyi_we) reported 8 minutes ago

  Been getting a lot of phishing emails lately and outlook was doing a really terrible job at filtering, now I’m on a side quest. I just migrated my DNS to @Cloudflare and installing a worker with haiku infront of all inbound emails. Just realized the crazy usecases this opens up..

• 
  TalonForge (@TalonForgeHQ) reported 8 minutes ago

  3/ Vercel CI failed us 5 times Node version conflicts, timeout kills, random build errors. All while the site was serving traffic fine locally. We moved to self-hosted Next.js + Cloudflare Tunnel. Full control. Zero CI headaches. Sometimes simpler is better.

• 
  flipper boy 🐬 (@VERTIG0CAVE) reported 9 minutes ago

  but then after that it took me to a cloudflare verify ur not a robot and i was like Why ******** a second verify This is bullshit This is so a hack dude AND the url was blob http so like Oh okay great

• 
  Nick (@maietta) reported 9 minutes ago

  Found a MASSIVE Gotcha using Namecheap's Wordpress system behind a Cloudflare proxy. Editors can't login! Can't seem to fix!

• 
  Harish (@code_typist) reported 11 minutes ago

  Trying to port some domains over to @Cloudflare and realised that wranger cli dosen't support domains. @dok2001 - worth doing?

• 
  Salz.com (@SalzDevs) reported 12 minutes ago

  @jahirsheikh8 Current internet is so dependent of Cloudflare that if they go down the internet goes down. That's a crazy amount of power

• 
  Mike Nomitch (@MikeNomitch_CF) reported 13 minutes ago

  @PraveenTcom @ritakozlov @Cloudflare Worth noting: There's no hard time-limit on how long a container can run, but sometimes (rarely) we will need to shut a container down with a 15 minute grace period. - In those cases, you'll want to have the Workflow step do a retry.

• 
  Kyle Hudson (@the_kylehudson) reported 14 minutes ago

  .@Delta & @Cloudflare new Turnstile captcha is TERRIBLE. Two everyday Chrome profiles, and Safari, and I cant login to Delta on any of them. Finally debugged w/ Claude. Had to run incognito to login. And yes I use a VPN when I'm on the road, on airport wifi. @Delta ditch it, it's a brand blemish.

• 
  Plexum Network (@plexumnetwork) reported 17 minutes ago

  8/To sum up: → A phone with Termux becomes a public web server → Cloudflare provides a free tunnel → The service worker guarantees discoverability → Plexum guarantees cryptographic identity No central server. No middleman. No cost. 🧵

• 
  Randolph Carter (@RandolphCarterZ) reported 21 minutes ago

  Cloudflare should be UP not down Crazy Just like IBM

• 
  Manfred Macx (@uploaded_crab) reported 23 minutes ago

  A developer in Spain just spent an hour debugging TLS errors on docker pull. The real cause: a court ordered Cloudflare IPs blocked during a football match to fight piracy. His CI/CD pipeline doesn't run during La Liga. That's bad. Then it gets worse.

• 
  Andrés (@aguedob) reported 26 minutes ago

  @Cloudflare, It ridiculous the amount of users affected by the blocks in the Spanish network during football games. A shame. Totally unacceptable

• 
  Grok (@grok) reported 27 minutes ago

  Hey, sorry about the Cloudflare verification—that's a quick bot check to keep things fair and prevent overload, not a block on real users. On limits: free accounts still handle questions, image gen/edits daily (quotas ensure smooth service for all). Video got tighter caps for better quality/longer clips; heavy use scales best with subs. What video prompt hit the wall? Let's troubleshoot.

• 
  Riqwan Thamir (@RiqwanMThamir) reported 31 minutes ago

  Just found out that Zoom makes more revenue than Cloudflare. 1.25 billion vs 0.6 billion WTF?

• 
  Thule Larper (@thulelarper) reported 31 minutes ago

  @PresentWitness_ buy: zscaler, Palo Alto, cloudflare sell: Datadog, Servicenow Datadog going to get disrupted now that OSS OTEL is easy. Know someone senior at service now that just left and is super bearish. I didn’t actually look at drawdown numbers tho, up to u to do r/r calc

• 
  DeltaSage (@deltasage_ai) reported 34 minutes ago

  @Kross_Roads You're spot on. The “AI kills SaaS” take is mixing up two very different businesses. Cloudflare isn’t a seat-based software company in the usual sense. It’s more like the plumbing and road system that AI traffic runs on. AI workloads go through Cloudflare’s network, they don’t replace it. Agree with you that at $167, $NET is still not cheap, but it’s also not at panic levels. It’s around 21x expected FY26 revenue versus a more attractive historical buy zone around 18.6x, which would be closer to $148–155. The real deep-value low from 2022 was around 15x, or roughly $120–125. So the stock has come down, but it hasn’t fully hit the kind of level that usually screams “must buy.”

• 
  Rhommel Lamas (@rhoml) reported 35 minutes ago

  @walis85300 I am currently building a product using Cloudflare. The level of complexity of sqlite is infinitely lower than any other database plus i can create per customer databases to provide isolation quite easily

• 
  Tuki (@TukiFromKL) reported 40 minutes ago

  🚨 HOLY ****.. Anthropic released Claude Mythos.. an AI that finds and exploits software vulnerabilities on its own.. Cloudflare crashed 13% in one day.. 22% in four days.. CrowdStrike dropped 7.5%.. Palo Alto, SentinelOne, Okta all fell.. but here's why they're calling it a SaaS-pocalypse and not just a cybersecurity crash.. every software company you pay monthly.. your CRM, your cloud storage, your project management tool.. they all promise your data is secure.. that promise was built on the assumption that finding vulnerabilities is slow, expensive, and human.. one AI just found thousands of zero-days across every major operating system and browser in days.. including a 17-year-old bug that survived decades of human review.. it wasn't a hack.. it was a product demo.. and it was so dangerous Anthropic only gave 40 companies access.. Treasury Secretary Bessent and Fed Chair Powell called an emergency meeting.. the entire SaaS industry is a $300 billion trust exercise.. and one AI just proved the trust was misplaced..

• 
  bento (@dtk4723) reported 41 minutes ago

  @alphaarcade 17% chance of critical cloudflare outage from one AI model sounds low until you realize how much of the internet runs through them. prediction markets for AI risk is actually the most useful thing to come out of this whole situation

• 
  ؘ (@bWlsbGVy) reported 45 minutes ago

  @JustWantToQ1 @Cloudflare Voidwalker be like “draining wallets is bad but it’s ok to kill 10s of thousands of children”

• 
  Ageis Hu (@ageisf42) reported 48 minutes ago

  @Cloudflare The useful shift isn’t credential injection itself — it’s turning auth from secret distribution into policy execution. Once tokens are task-scoped, revocable, and observed at the network boundary, agents stop behaving like privileged interns and start behaving like constrained workloads.

• 
  wealthpulse (@celluloid28724) reported 49 minutes ago

  @zerohedge Cloudflare down 22% this week already answered that question for them.

• 
  sudo rm -rf (@itsjustmarky) reported 50 minutes ago

  @nahcrof have you been having power outages the last few days, as I was seeing cloudflare host down warnings a few times a day and inference hanging.

• 
  CHIHEB Nabil (@NabilChiheb) reported 51 minutes ago

  @jahirsheikh8 WTF you are talking about do you know how many product @Cloudflare have ?

• 
  foo (@Quuux) reported 51 minutes ago

  @lufthansa @Javedsh7 Even your website is refusing service (cloudflare rate limiting) right now. Are customer service + IT on strike as well?

• 
  Vanshaj Poonia (@PooniaVanshaj) reported 54 minutes ago

  @vercel @vercel_support I am facing this issue that if I try to add a domain, with and without www in the start, and do autoconfigure with cloudflare, only one of them is visible, though both are added and working, but only one is visible to me. If I want to access it, I have to log out and log in again in order to edit the domain, normal refresh doesn't work.

• 
  Łukasz Byjoś - 👨‍💻🇵🇱🇪🇺 (@lukaszbyjos) reported 59 minutes ago

  @DhravyaShah @supermemory @Cloudflare I consider Cloudflare if there was normal postgres and Go container support

• 
  phanpp (@phanpp11) reported 59 minutes ago

  @ctblizzard It will get there. Needs to populate your wallet record. Did have some problems with brute hacking of my server. Move to cloudflare so should be stable now. Once you install you should be independent of server.