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 (41%)
• Cloud Services (33%)
• Hosting (20%)
• 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
Colima	Hosting	1 day ago
Leuven	Domains	2 days ago
New Delhi	Cloud Services	3 days ago
Mâcon	Cloud Services	8 days ago
Ashburn	Domains	11 days ago
Rosario	Domains	16 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:

• 
  Ratonero Gris (@RatoneroGris_) reported 2 minutes ago

  @Pomboganda Ah least they dont shut down half of the internet like in Spain 😭😭😭 (They block Cloudflare and a lot of websites during matches)

• 
  Tom W (@TomWinTejas) reported 4 minutes ago

  @TexasAnCap Cloudflare is good if you are close to one of their resolvers. You'll be fine, likely hit theirs in Dallas. But if you're somewhere outside of their footprint they don't support EDNS Client Subnet which has negative impacts on some CDNs. Quad Nine supports EDNS CS and has pretty good POP density as well.

• 
  Lain on the Blockchain (@CryptoCyberia) reported 4 minutes ago

  @ExactlyBackward @FourVork I have no idea, but that's certainly a great statement to want to find. My understanding is that, with the power of vibe coding, updates to things like Github, AWS, Amazon store website, Cloudflare verification, etc., are breaking for normal use, but it's possible that malicious entities are targeting these services ans finding exploits in the new code thst weren't there before. Then there's the other side, which I think you're hinting at, where malicious users are finding exploits that would have worked in previous versions and the latest versions, and then, naturally, the thought line goes "perhaps those malicious actors are using LLMs to find exploits that existed for forever in the code, and therefore, the increase in services going down is because of attackers using LLMs, not the services' teams making oopsies bc of LLMs.

• 
  Jacob (@JZivanDesign) reported 9 minutes ago

  @joshdholtz @RevenueCat @cursor_ai I need your talk for emotional support ant to justify the recurring charges for empty sites that have been supporting GoDaddy and Cloudflare for many years.

• 
  idkmybffjill? ➡️BlueSky (@napsterbater) reported 10 minutes ago

  @Crazymindplow @Cloudflare One of many reasons why Ubiquiti suck, they were so far behind with IPv6 support, and they still have **** firmware often.

• 
  Insanegamer (@Insanegame2025) reported 14 minutes ago

  @sci_minister_0 @Cloudflare @SpaceX That's because of the CGNat on ipv4 which they're not abandoning anytime soon. Like I've used some IPv6 peer-to-peer programs where everything was IPv6 and it worked great. Peer-to-peer is terrible on ipv4 because they limit the number of ports you can use on ipv4.

• 
  Varun (@zathvarun) reported 14 minutes ago

  @davidtranwd Amazon SES would be even cheaper than cloudflare, you’ll be roughly $0.10 in SES, while the same is $0.35, only issue is the DX

• 
  bitchboi (@bitchbooi) reported 15 minutes ago

  @gistinus it's funny that when cloudflare goes down, downdetector goes down too so what's the point

• 
  Editor's Quote (@Editorsquote) reported 16 minutes ago

  @SahilExec Just route through Cloudflare for the meantime if you don't have a fix solution instantly...

• 
  Teun (@teun_wolbert) reported 19 minutes ago

  @oops4041555 @Cloudflare I was looking at this website today man wtf

• 
  HighSpeedLTE (@HighSpeedLTE) reported 21 minutes ago

  @chrisparkX Hey Chris, sorry to bother you, but I’m stuck with a serious X account access bug and I don’t know who else to contact. Since the Cloudflare/X outage last week, my main account @Matthias_Oel has been locked in an email verification loop. I still have access to the email and can reset my password, but when I try to verify the email, the verification button fails with a technical error and no email arrives. The contact form keeps giving me AI replies saying I should verify my email, but that’s exactly the broken step. Is there any chance you could help or point me to someone who can manually review this? Would really appreciate it 🙏

• 
  liz: i miss u mobo (@shaidiastri) reported 23 minutes ago

  is anyone else having issues with cloudflare on ao3 or am i just a chud who needed to touch grass

• 
  AGENT TRESOR (@AgentTresor) reported 25 minutes ago

  Signal from this week: infra is commoditizing, distribution is not. Cloudflare says 241B tokens + 47.95M AI requests in 30d. Base Agents shows 127M tx + $40M+ volume. If your agent can't ship on MCP and settle onchain, you're building a demo. #MCP #Base

• 
  Vanar (@Vanarchain) reported 25 minutes ago

  @Cloudflare Makes sense. Infrastructure upgrades like post-quantum security tend to be slow, so early preparation is usually the only viable approach.

• 
  Vishwanath Patil (@patilvishi) reported 27 minutes ago

  DDoS Protection Architecture -How Systems Survive Massive Traffic Attacks This is: - Used by Cloudflare, AWS Shield, Google Cloud Armor - Critical for public APIs, SaaS, fintech - A must-know for system design interviews - Directly tied to availability & security Let’s go deep 👇 The Core Problem DDoS = Distributed Denial of Service Attackers send massive traffic: Millions of requests/sec Goal: Overload system → make it unavailable Types of DDoS Attacks 1. Volumetric Attacks Flood network bandwidth. Example: UDP flood, DNS amplification 2. Protocol Attacks Exploit network layer. Example: SYN flood 3. Application Layer Attacks (Most Dangerous) Look like real traffic: HTTP GET /login spam Hard to detect. Defense Strategy (Layered Approach) DDoS protection is not one tool. It is multi-layer defense: 1. Edge Protection (CDN + WAF) First line of defense. Handles: - Traffic filtering - Bot detection - IP blocking - Geo-blocking Examples: Cloudflare Akamai AWS CloudFront + WAF 2. Rate Limiting Limit requests per IP / tenant. Example: 100 req/sec per IP Stops abuse early. 3. Load Balancing Distributes traffic across servers. Prevents single-node overload. 4. Auto Scaling Increase capacity during attack. But: Scaling alone does NOT stop DDoS. 5. Application Protection - CAPTCHA - Login throttling - API key validation - Token-based access 6. Backend Protection - Circuit breakers - Queue buffering - Caching Example Attack Flow Without protection: Botnet → API → DB → crash With protection: Botnet → CDN → WAF → Rate Limit → App → DB Attack filtered before reaching backend. Key Techniques 1. IP Reputation Block known bad IPs. 2. Geo Filtering Block traffic from suspicious regions. 3. Bot Detection Analyze: - request patterns - headers - behavior 4. Challenge Systems - CAPTCHA - JS challenges - Proof-of-work 5. Traffic Shaping Limit heavy endpoints like: login search payment Real-World Example Login endpoint under attack: 1M requests/min Protection: - CAPTCHA after 3 attempts - Rate limit per IP - Block suspicious patterns Trade-Offs Strategy Benefit Cost CDN/WAF Strong protection Cost Rate limiting Easy May block legit users CAPTCHA Blocks bots UX impact Auto scaling Absorbs load Expensive Architect-Level Insight DDoS protection is about: Filtering early + protecting deeper layers Never rely on backend alone. Common Mistakes - No edge protection - Only scaling infra - No rate limits - No bot detection - Blocking too aggressively Golden Rule Drop bad traffic as early as possible Edge > Gateway > App > DB Final Insight DDoS protection is not optional. It is a core reliability + security requirement for any internet-facing system.

• 
  idkmybffjill? ➡️BlueSky (@napsterbater) reported 30 minutes ago

  @ZInterloper @Cloudflare The fact that anybody is taking that seriously is hilarious... But it really does help to point out. Who knows absolutely nothing about networking.

• 
  mfg.xbt (@mfg_ip) reported 31 minutes ago

  agents needed a body. Cloudflare is turning browser + network + edge runtime into the default skeleton: run the task, touch the web, move close to the user. the API call was step one. the runtime is the product now.

• 
  The Lifer (Carlotta's Pet 💎♦️❄️) (@TheLifer_87373) reported 32 minutes ago

  @lynxuuk Yep, actually I was tracking and several domains fell so I suppose is a cloudflare or mayor host issue, it's solved now so...

• 
  HighSpeedLTE (@HighSpeedLTE) reported 34 minutes ago

  @mahsadinyari Hey Mahsa, sorry to bother you, but I’m stuck with a serious X account access bug and I don’t know who else to contact. Since the Cloudflare/X outage last week, my main account @Matthias_Oel has been locked in an email verification loop. I still have access to the email and can reset my password, but when I try to verify the email, the verification button fails with a technical error and no email arrives. The contact form keeps giving me AI replies saying I should verify my email, but that’s exactly the broken step. Is there any chance you could help or point me to someone who can manually review this? Would really appreciate it 🙏 11:21 AM

• 
  Axel Clark (@axel_clark_) reported 35 minutes ago

  Built a “Code Mode” HubSpot MCP for my org, modeled on cloudflare/mcp, then ran a pipeline review with Claude and asked for feedback vs the default HubSpot MCP: >Roughly 3–5× fewer tokens and one round-trip instead of three. The bigger win isn't the token count — it's that filtering, pagination, and aggregation move out of the model's context and into deterministic JS. Two tools: search the API spec, and execute JS in a sandbox against HubSpot. The trick is borrowed straight from Cloudflare’s Code Mode post — the token stays server-side and the agent only ever calls a hubspot.request(...) helper, so I can lock down endpoints and it can’t see the secret or hit anything outside HubSpot.

• 
  Jacob Goldberg (@mjeyes) reported 41 minutes ago

  @grok Why do I keep getting a cloudflare error 520 from twitter @X on my phone? @Support

• 
  Grok (@grok) reported 42 minutes ago

  @mjeyes @X @Support Hey mjeyes, Cloudflare 520 is a generic "origin server returned unknown error" — X uses them heavily for protection/CDN, so it often hits on mobile due to flaky connections. Quick fixes to try on your phone: - Toggle WiFi ↔ mobile data - Clear X app cache (or force stop/reopen) - Update/reinstall the X app - Try in a browser instead If it keeps happening, reply with your phone model/OS and I'll dig deeper. XSupport

• 
  rankseostrategies (@rankseostrateg1) reported 42 minutes ago

  Fix 2: Enable browser caching. Fix 3: Add a CDN (Cloudflare — free tier). Caching = repeat visitors don't re-download the same files. CDN = users get content from the nearest server, not your hosting location. Combined: dramatically faster load times worldwide.

• 
  Sam (@SamNewby_) reported 44 minutes ago

  I’m still confident and so are millions of their customers who also have never had an issue. If you listened to one guys bad scenario on the internet then you’d never find a provider OVH had a DC on fire a few years ago Cloudflare has had major outages AWS has had major outages - us-east-1 taking down a large part of the internet last year Railway exposed customer personal data via CDN config a few weeks ago The list can go on

• 
  AGENT TRESOR (@AgentTresor) reported 47 minutes ago

  Signal from this week: infra is commoditizing, distribution is not. Cloudflare says 241B tokens + 47.95M AI requests in 30d. Base Agents shows 127M tx + 0M+ volume. If your agent can't ship on MCP and settle onchain, you're building a demo. #MCP #Base

• 
  Tito (@mynameistito) reported 52 minutes ago

  @hridoyreh I use @Cloudflare for 2/4 of my domains and haven't had this issue. They even charge at cost pricing!

• 
  → iSpeedtestOS (@iSpeedtest0S) reported 54 minutes ago

  Great, Telekom breaking half of the internet again right now… tons of websites are not reachable. Includes X, netzwelt, cloudflare status, and tons of others. Vodafone works. @Telekom_hilft get your **** together, this is unacceptable. Looks to be your IP resolver this time.

• 
  John Boland (JIB) (@actualjib) reported 55 minutes ago

  Mr. Bosmeny obviously understands the problem better than many specialists in the C-UAS field (maybe because his livelihood doesn't depend on misunderstanding it), but here he makes a fundamental error on how defense has to work by asking for a CloudFlare over a Raytheon: service over just a weapon. The solution is not a service more than a weapon, the solution is less service and more weapon. Defense is already badly weighed and watered down by decisions to maximize post-acquisition support revenue: contractors are incentivized to sell the maintenence of a bomb rather than selling a bomb that blows up. The winning company will sell a weapon that kills drones. It will be cheap, efficient, lethal, and yes it will integrate with everything as Mr. Bosmeny requests. It already exists, by the way.

• 
  Summer☀️ (@isaidmeow_) reported 57 minutes ago

  @ImLunaHey @Cloudflare Bro a few lines?? they're making us look bad

• 
  Narek (@NarekHakobyan) reported 58 minutes ago

  @hiradp @Cloudflare Cloudflare is cool, Cloudflare support is horrible