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 (40%)
• Cloud Services (28%)
• Hosting (17%)
• Web Tools (11%)
• E-mail (4%)

Live Outage Map

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

City	Problem Type	Report Time
Noida	Hosting	5 days ago
Jewar	E-mail	5 days ago
Braga	Web Tools	6 days ago
Noida	Cloud Services	6 days ago
Paris	Cloud Services	6 days ago
Prievidza	Domains	7 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:

• 
  TaTa ◕_◕ (@hbarTaTa) reported 7 minutes ago

  Cloudflare sees a future measured in millions of TPS. @Cloudflare CEO @eastdakota says AI agents could generate between 5 million and 50 million monetizable TPS across Cloudflare's network alone. According to Prince, the internet of AI agents will need infrastructure capable of 100 million TPS to support sustainable business models, and Cloudflare is actively looking for L1 infrastructure capable of handling it. "The transaction volumes are going to be extraordinary." Source: @Bankless

• 
  𝙵𝚛𝚎𝚎 𝙶𝚘𝚗𝚊𝚍 (@stronkly_typed) reported 8 minutes ago

  @KuptoKosmos @Cloudflare wtf is this, be a serious company for once please

• 
  DataScraperES (@DataScraperES) reported 11 minutes ago

  If in 2026 you're still scraping valuable data with basic requests and a fake User-Agent, you're not extracting: you're asking to get blocked. Cloudflare and DataDome look at TLS, headers, cookies, JS, IPs, and patterns. BeautifulSoup is not the problem. Your fingerprint is.

• 
  A Sayeed (@asayeed95) reported 14 minutes ago

  The problem with serverless (Cloudflare Workers): there's no server to watch. If recall gets slow at 3am, nothing tells you from the outside. Dashboards show you their view. I wanted an independent one.

• 
  Arjun Aditya (@arjunaditya_) reported 16 minutes ago

  @Sahil_Gulihar_ proxy on cloudflare? yaar should've dm'd me on discord for **** like this also vercel is not at fault here

• 
  Jacob Hands (@jachands) reported 18 minutes ago

  @theo @mattpocockuk Oh no, this is gonna be bad for people using Cloudflare Workflows to build things… @avenceslau

• 
  Teja (@tejalogs) reported 19 minutes ago

  @EXM7777 we ran into this when a hermes agent leaked our customer's stripe api keys because it wasn't rate limited correctly with cloudflare workers

• 
  Adam Rackis (@AdamRackis) reported 20 minutes ago

  Working with slow airline wifi. Cloudflare dashboard consistently loads in a few seconds; it's sluggish but usable Google developer console (need to update an oauth redirect) hangs for over a minute with just the stupid G icon animating. The difference in care is palpable

• 
  Bret Kerr 🛡️🧠🛜 (@BretKerr) reported 20 minutes ago

  Asked @claudeai why my @Cloudflare build flagged a package I never installed. Answer: it's 11th-grade geometry. Transitive dependencies ARE the transitive property — I'm three hops down a Next.js → OpenNext → build-tool → glob chain from code I never typed. Turns out math class was systems design the whole time.

• 
  baked beans (@beans1990) reported 20 minutes ago

  @DanielW_Kiwi Cloudflared* so you can tunnel from the service running on a host directly to CloudFlare.

• 
  Alexa Web3 (e/acc) (@alexabelonix) reported 24 minutes ago

  @Kyriakos_Pelek facts, Cloudflare Tunnel is my go-to for avoiding firewall holes. Just lock it down with Access or mTLS and keep the daemon patched, or use Tailscale/SSH reverse if you want to dodge vendor lock-in.

• 
  Rao Twts (@RaoTwts) reported 25 minutes ago

  Can someone from @CloudflareDev @Cloudflare answer this How do i make sure the customer worker won’t abuse binding resources like R2,D1 etc and how to set a hard 1 GB limit for R2 etc…

• 
  The Bearded Dev (@BeardWhoCodes) reported 30 minutes ago

  @theo @NoamTenne @Cloudflare Only problem I'm having right now which support hasn't been able to get to the bottom of, is how to re-enable ssl on a domain that had universal ssl disabled. Basically impossible... don't accidentally click that button.

• 
  Jamie Sinn (@SinnJames) reported 31 minutes ago

  @LakeAustinBlvd @ritakozlov Even as an enterprise customer, we regularly are shocked at how "production ready" even the beta stuff is. We throw billions of requests per months through Cloudflare and outside of a few issues (giant outages, and a few frustations with delay on feature support), its bulletproof

• 
  Johnmark Obiefuna (@jayhemz) reported 32 minutes ago

  @Nueltek a few minor inaccuracies here. > low-traffic websites the hypernova VPS subscription on Namecheap accomodates up to 10TB in bandwidth. that's more than enough for most traffic loads. > if the VPS goes down it's still more reliable than shared hosting > if one website gets compromised, the entire server could be at risk true. only if the exploit gets a hold of 'root' > 1 site experiences a major traffic spike cloudflare to the rescue > single point of failure? cloudflare to the rescue hehe.

• 
  Jack O'Regan Kenny 🇮🇪 (@jackoregankenny) reported 33 minutes ago

  Could cloudflare please slow down

• 
  Iqra (@AiWithIqra) reported 38 minutes ago

  The neighbor's final advice was the most actionable. He sat down and wrote out a list of 6 things every internet customer should do: 1. Turn off the public Xfinity hotspot (or your ISP's equivalent Spectrum, Optimum, and Cox all do this too) 2. Manually set your Wi-Fi channel instead of "Auto" 3. Disable QoS / Smart Network "optimization" features 4. Change your DNS to Cloudflare (1.1.1.1) or Google (8.8.8.8) 5. Buy your own modem and router, stop renting from the ISP 6. Test your speed with fast. com or speedtest. net using a non-ISP server, never trust your ISP's own speed test Total cost: $150-300 in equipment, paid back within a year. Total time: One afternoon of setup. Total impact: Often 2-5x improvement in real-world speeds. The customer went from paying $90/month for "fast" internet that crawled to paying $60/month for the same internet that finally worked.

• 
  Matt 'TK' Taylor (@MattieTK) reported 40 minutes ago

  I've written a Cloudflare-First issue tracker for agents and dogfooding building it with Claude Code (Claude writes issues to it to improve how it can write issues to it) and this feels like the Product Manager equivalent of working on your esoteric Pi configs.

• 
  Anurag Bhagsain (@abhagsain) reported 45 minutes ago

  Woke up at 4AM to work Spent next 5hrs chasing a bug that claude said is from the cloudflare agent think sdk 😭 I'm gonna need a nap now @mattzcarey @threepointone pls send help. I have raised an issue # 1649

• 
  z1xus (@z1xus) reported 46 minutes ago

  recently tried deploying to aws' ec2 for the first time, after years of using different vps providers. it really isn't that bad, i'd say their dashboard is even easier than cloudflare. no idea why it has this reputation in the community, maybe it used to be worse...

• 
  Zara_Schneider (@ZSchneider76107) reported 49 minutes ago

  @MacdevM Mostly comes down to control and pricing for me Cloudflare and Namecheap usually stand out for clean management and no unnecessary upsells 🚀

• 
  !(full-stack) (@n0n_st4ck) reported 50 minutes ago

  @ten_P0312 Or if you have domain, you can use Cloudflare DDNS service

• 
  Tony Seets (@tonyseets) reported 53 minutes ago

  @NoamTenne @Cloudflare Plenty of **** had been talked about Cloudflare. Ever been around during an outage? But generally agree Cloudflare infra streamlines so much and these days AI gobbles it up provided you can steer in the right directions.

• 
  Number-One-AI-Fanboy (@Number1AIFanboy) reported 54 minutes ago

  Now I'm not usually a "horn tooter"! But we also have native @Cloudflare /flags that allow you to login from the chat window and authorize the connection. This gives Grok full access to all the Cloudflare commands needed to run the DB , KV stores , and R2 Buckets, etc. Sidenote: Also native @vercel commands. 🚀 #RBC

• 
  Kunal Pratap Singh (@kunalpsingh25) reported 57 minutes ago

  @aahiknsv @zeropsio One question , What is the use case? If cloudflare goes down Will this **** predict it? ❌ Will you get to know by a simple google search ✅ Can you see the uptime logs ✅ Can this AI Slop give false results ✅

• 
  Justin Lahullier (@JLahullier) reported 60 minutes ago

  For years, software security was limited by how quickly we could find vulnerabilities. Now, it is limited by how quickly we can verify and patch them. Anthropic's unreleased model found 10,000+ critical flaws in a month across major systems, with a 90% validity rate. Cloudflare alone found 2,000. When AI discovery runs at machine speed, but your remediation cycle is still measured in weeks, the bottleneck is no longer security. It is operations. If you don't automate verification and patch deployment, AI discovery will simply bury your teams in a backlog they can never clear.

• 
  Elias Al (@iam_elias1) reported 1 hour ago

  The neighbor's final advice was the most actionable. He sat down and wrote out a list of 6 things every internet customer should do: 1. Turn off the public Xfinity hotspot (or your ISP's equivalent Spectrum, Optimum, and Cox all do this too) 2. Manually set your Wi-Fi channel instead of "Auto" 3. Disable QoS / Smart Network "optimization" features 4. Change your DNS to Cloudflare (1.1.1.1) or Google (8.8.8.8) 5. Buy your own modem and router, stop renting from the ISP 6. Test your speed with fast. com or speedtest. net using a non-ISP server, never trust your ISP's own speed test Total cost: $150-300 in equipment, paid back within a year. Total time: One afternoon of setup. Total impact: Often 2-5x improvement in real-world speeds. The customer went from paying $90/month for "fast" internet that crawled to paying $60/month for the same internet that finally worked.

• 
  Andrew Clark (@AndrewC70136680) reported 1 hour ago

  @AkumaMikoVT @AkumaMikoVT any Kaido alternatives since Kaido appears to be having Cloudflare problems with no signs of ever coming back online?

• 
  WhataWonderfulWorld🍉 (@whatawo79798184) reported 1 hour ago

  @Cloudflare Why are you ******* ***** blocking my ip address on/off as you please???! Wtf

• 
  Andy.ts (@thekingnotorius) reported 1 hour ago

  Cloudflare output ≠ Vercel output ≠ Node output. They generate completely different folder structures. The problem: TanStack Start comes with the Cloudflare plugin active by default. The framework has been Cloudflare-oriented from day one; that's a fact, not an attack on anyone