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.

• Cloud Services (43%)
• Domains (21%)
• Hosting (21%)
• E-mail (7%)
• Web Tools (7%)

Live Outage Map

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

City	Problem Type	Report Time
New York City	Hosting	15 hours ago
New York City	Cloud Services	19 days ago
Los Angeles	Cloud Services	21 days ago
Paris	Cloud Services	1 month ago
New York City	Hosting	1 month ago
Manchester	Domains	2 months 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:

• 
  developing valhalla (@valhalla_dev) reported 5 minutes ago

  Notably, most technology that has been lifechanging and revolutionary is technology that the average person does not at all care about. The average person doesn't care about Linux, or Docker, or the HTTP protocol, or virtual hosting, or React. They don't have to. They like what comes downstream of that technological progress. So the argument that "AI doesn't matter to most people" is moot, because the downstream effects of AI do matter. Yes, AI People need to stop pretending like AI is going to steal everyone's jobs yesterday, yes they need to stop shoehorning AI into places that don't matter. But the answer to "why should the average person give a damn about AI" is generally "they probably shouldn't." The market is going to get kinda crazy, there is going to be a lot of tumult around data centers and inference buildout, and a lot of new consumer facing and business facing tech is going to be released at an accelerating rate. But the average person shouldn't really have to care about AI, any more than they care about what a VPS is or how CloudFlare Pages work.

• 
  Odogwu engineer exploring ML (@slycreator1) reported 6 minutes ago

  Been doing software for years, but it was just last quarter that I opened a Cloudflare account because of this issue. I insisted on using Cloudflare's free email routing cause it’s only accessible with VPN

• 
  YanXbt (@IBuzovskyi) reported 11 minutes ago

  HERMES AGENT HAS AN AUTONOMOUS SKILL MANAGER RUNNING IN THE BACKGROUND. IT PRUNES WHAT YOU DON'T USE. IT ARCHIVES WHAT'S STALE. IT'S CALLED THE CURATOR. every time your agent solves a new problem and saves a skill, that skill lands in ~/.hermes/skills/. month 1: 15 skills. manageable. month 3: 47 skills. some are near-duplicates. month 6: 90+ skills. half are stale. catalog bloated. tokens wasted every turn loading skill headers the agent never calls. the Curator cleans this up automatically. HOW IT WORKS: tracks every skill: how often viewed, used, patched. 30 days unused → marked stale. 90 days unused → moved to archive. archived skills go to ~/.hermes/skills/.archive/ never deleted. always recoverable. runs every 7 days. enabled out of the box. WHAT IT TOUCHES: agent-created skills: full lifecycle. stale, archive, consolidate, patch. bundled built-in skills: archived after 90 days of non-use (on by default). hub-installed skills (agentskills. io): NEVER touched. off-limits. WHAT IT DOES NOT TOUCH: memory files. USER.md and MEMORY.md are not managed by the Curator. people confuse this constantly. Curator = skills only. PIN YOUR CRITICAL SKILLS: hermes curator pin deploy/cloudflare-pages pinned = off-limits. no stale. no archive. no consolidation. skill_manage refuses to delete it. patches still go through (agent can improve content). only agent-created skills can be pinned. skills in any cron job's skills: list are also protected without explicit pin. COMMANDS: hermes curator status (state + 5 least-used skills) hermes curator run (manual run now) hermes curator run --dry-run (preview without changes) hermes curator pause (stop auto-maintenance) hermes curator resume (restart) hermes curator pin [name] (protect) hermes curator unpin [name] (remove protection) hermes curator backup (manual snapshot) hermes curator restore [name] (recover archived skill) hermes curator rollback (revert last run) also available as /curator inside any session. LLM CONSOLIDATION (off by default): opt-in auxiliary model pass that reviews your skills for overlaps and proposes merges. curator: llm_review: true WARNING: run --dry-run first. always. auto-consolidation can merge skills you want to keep separate. comment CURATOR and I'll send you the full checkup guide with 5 use cases and config templates for solo dev, multi-project, and team setups.

• 
  njoii.eth (@DanSFV) reported 14 minutes ago

  Blackwell build update: Today I restored the watchdog layer and got production back to a fully healthy state. The watchdog now continuously checks: • Bridge • Runner A • Runner B • Cloudflare • duplicate processes / port conflicts And the important part: it already caught a real Runner B health failure, restarted it automatically, and returned the system to healthy without me touching anything. That’s exactly why I’m building this layer. A trading system shouldn’t depend on me noticing a service died. The infrastructure should detect the failure, recover, verify the state, and keep the rest of the system isolated. Current state: Bridge healthy Runner A healthy + flat Runner B healthy + flat Public webhook reachable Single listeners only Watchdog active Production is frozen now and I’m waiting for the next natural Volume Profile signal to validate the full entry + exit path. The goal is getting closer to what I want Blackwell to become: not just automated trading. Automated recovery too.

• 
  Andrew Rulnick (@MickeySteamboat) reported 16 minutes ago

  @sierracatalina going to say there's a good chance it's hardware related sometimes it can be internet provider based, I had cloudflare challenging my nets and it was causing a LOT of problems with GPT in general until I identified that was the issue. ) :

• 
  PangeaVPN (@PangeaVPN) reported 17 minutes ago

  @HorluwarhArmani @SirAlexanderrr if it's the ISP breaking that page rather than Cloudflare, you need something the network can't pick out. Pangea tries five stealth transports in order, VLESS+Reality first, so one being blocked isn't the end of it. windows/mac, ios via shadowrocket. five days free.

• 
  Fabrice NEYRET - pro (@FabriceNEYRET) reported 18 minutes ago

  @KaleyGoode cloudflare is ultra-annoying but usually you can log-in ! (from time to time you need to re-confirm identity once logged + and some days ago the site was down, though). BTW it seems that the unofficial plugin is (sometime?) incompatible with cloudflare.

• 
  Scroll Society (@maximcappaiv_) reported 19 minutes ago

  @Cloudflare @ATT @SacCountyCA is either switching their systems. I’m important here because the operating system is stationed down the street. Lots of guys want to plug in their own software but a good decision would be to improve the connection from static polls which were in the Eureka area of

• 
  Elijah 🌊 (@juiceboy_of_abj) reported 19 minutes ago

  @Ms_Ada6 Nahh is not network and cloudflare is not even the only one

• 
  Mildlysupreme (@mildlysupreme) reported 19 minutes ago

  @Namecheap @Cloudflare unfortunately namecheap's web app is so ridiculously slow on top of the price hikes.. this is going to take a while

• 
  Brad Gessler (@bradgessler) reported 22 minutes ago

  @crqrdotcom Putting your name down for cloudflare.

• 
  Hareesh Vemasani (@HVemasani) reported 24 minutes ago

  If you want to start a startup: Claude = coding. ($20/mo) Supabase = backend. (Free) Vercel = deploying. (Free) Namecheap = domain. ($12/yr) ProductBridge = customer support + feedback (Free) Stripe = payments. (2.9%/transaction) GitHub = version control. (Free) Resend = emails. (Free) Clerk = auth. (Free) Cloudflare = DNS. (Free) PostHog = analytics. (Free) Sentry = error tracking. (Free) Upstash = Redis. (Free) Pinecone = vector DB. (Free) Total monthly cost to run a startup: ~$20

• 
  Sid J (@SidJ79) reported 26 minutes ago

  Running one Chromium instance per agent is too expensive to scale. That's the problem it's solving. Cost, not speed. Cloudflare says they'll open source it.

• 
  Faye Xiao (@faye_xiao_) reported 29 minutes ago

  The spirit of Spirit just sold for $10 million Google is buying Spirit Airlines' data out of bankruptcy. Emails, internal communications, spreadsheets, bookings, frequent flyer and HR records, all de-identified, for $10 million. Judge Sean Lane rules on the sale Wednesday. The obvious read is that Google wants more data. But de-identified data doesn't work for advertising, since you can't target someone you can't name, and Google already sees more airfare information through Google Flights than Spirit ever generated internally. An airline that went under in May is also a strange place to look for pricing wisdom. What's worth buying is the internal material. The emails and the spreadsheets they reference record how work moved through the company: a question gets asked, a document gets built, a decision gets made, a system gets updated. Consumer text is everywhere, and records of how an organization actually functions are not, which is what you need if you want models that operate inside workflows rather than talk about them. Google's own statement uses the word enterprise, and the runner-up bid of $7.5 million came from Mercor, a company whose entire business is sourcing training data for AI labs. When the second bidder isn't another airline, the market has told you what was being priced. The strange part is that Google isn't short on this data at all. It runs Gmail and Workspace and sits on possibly the largest collection of business correspondence in the world, and it has promised enterprise customers it will not train on their content, which is not a promise it can quietly break. So it has the material and no permission to use it. What $10 million buys is clean title, a court approved dataset nobody can sue over, at a moment when everyone else is defending scraping claims. Dead companies can agree to things live ones can't. Any of this is worth paying for because the public supply is running down. Epoch AI's 2024 analysis put the stock of quality public human text at roughly 300 trillion tokens and projected it would be consumed between 2026 and 2032, a window that opens this year, and access has closed faster since than the arithmetic alone suggests. A census of the top 100,000 domains this July found 19.1% blocking at least one AI crawler, and in September Cloudflare begins blocking mixed use crawlers by default across its entire free tier. Private operational records are the obvious next reserve, and they are almost untouched. Epoch left them out of its estimate because private data is fragmented and legally too messy to use at scale, which is precisely the condition a bankruptcy court removes. That makes Wednesday's ruling more interesting than the sale. If it goes through, every bankruptcy from here has a new asset to offer, and the value will depend on how well documented the industry already is. A corner store has nothing worth buying, since you can watch how it works from the sidewalk. A hospital or a law firm is the opposite, because even with names removed the record shows how a case moves through the organization, who escalates what to whom, and which exceptions get made. That knowledge lives in internal systems and in people's heads and appears nowhere public. The catch is that the supply is biased toward failure, since no healthy company would sell its internal record, so every dataset that reaches the market comes from an operation that didn't work. That's useful for learning how a process runs, and much less useful for learning what good judgment looks like.

• 
  vivek (@vivek31337) reported 30 minutes ago

  Don’t spend on marketing tools. Do this for free Build a system where AI agents have access to PATs/APIs for GitHub, vercel , cloudflare , Render, Gumroad, Stripe, etc. Then keep them in a loop: Research → find ideas → generate keywords → save to file → build website/product → deploy → market → track sales → research again. ask agents to submit to google search console. Tell your hermes to note down things , for reports tasks / status/ needs ask chatgpt for telegram script so your agents will run automatically, even if they get stuck somewhere they will move on to another task. my agents have shipped 300+ websites/ pages / articles in 20+ languages in few hours , i can bet no human on earth can do that. if you need free tokens lmk.

• 
  Manish Malhotra (@algotradingdesk) reported 33 minutes ago

  Cloudflare is proving that cybersecurity is becoming an infrastructure business rather than simply a software category. The company recently crossed $2 billion in annual revenue, while its network spans more than 330 cities globally. But the more important number is the amount of internet traffic passing through infrastructure like this. As AI agents multiply, automated traffic and cyberattacks increase alongside legitimate users. That creates a structural shift. Every new AI agent becomes another potential customer. Every new attack becomes another reason to pay for protection. Cloudflare isn't merely selling security. It is positioning itself between the internet and the companies trying to operate on it. In the AI era, the companies controlling digital traffic may become as strategically important as the companies generating it.

• 
  Drayken (@KGDrayken) reported 35 minutes ago

  @mdong1909 Cloudflare owns both VoidZero and Astro now - can you make it a priority to have full support of Astro for Oxfmt and Oxlint now so monorepos don't have to fragment into linting & formatting fallbacks? IIRC plan was Q2.

• 
  Prince Batubo (@Iamprincedawari) reported 40 minutes ago

  @juiceboy_of_abj Nah... Mine is the other way round, I can't use MTN to sign-up or login on websites that's Cloudflare gated

• 
  Matt Smith (@msmith508) reported 41 minutes ago

  My router started having issues downloading a firehol ip block list…problem appears to be compound: they now block forcing redirect to but even after importing trusted root certs, it still fails, due to cloudflare most likely…my current workaround is dl from github

• 
  Uben (@Ubendev) reported 43 minutes ago

  @aiwithadb just shipped the version 4 of my CRM extension. mostly Cloudflare worker issues fixes.

• 
  FYAZ (@OgFyaz) reported 43 minutes ago

  🚨 A Spectre attack leaked a JWT from a co-located Cloudflare Worker. Researchers demonstrated the attack in Cloudflare’s production environment at up to 12 bits/second, 360× faster than the 2021 result. No customer data was accessed. Cloudflare says the attack is now mitigated. Thoughts?

• 
  Ossy Vincent (@ossynoya) reported 45 minutes ago

  To fix this issue, at least what I did then, either of these methods works perfectly. 1: Download Cloudflare WARP on your phone or laptop. It works like a lightweight VPN by routing your internet traffic through Cloudflare, so you should be good to go. 2: Go to your Airtel Wi-Fi login page, open Settings, and change the Primary DNS to 1.1.1.1 and Secondary DNS to 1.0.0.1 I used these settings on my Airtel ODU and never had any issues accessing websites.

• 
  Gneaus (@legibitiquas) reported 47 minutes ago

  @juiceboy_of_abj Not just cloudflare. There was a time Outlook looses connection when I switch to Airtel, but other sites work. I had to switch back to MTN, they have fixed that now. Airtel have serious DNS issues, I noticed I can't open some sites, gives me DNS error

• 
  Aaron Makelky (@theaaron) reported 48 minutes ago

  @nickvasiles just point them at a private github repo or cloudflare artifact where your skills live installing them natively across different agents is a waste of time and they never stay synced to the system of record version

• 
  David Aronchick (@aronchick) reported 49 minutes ago

  @QuinnyPig @Cloudflare @vercel When do **** posts get their arn

• 
  IND_is_Here (@IND_is_Here) reported 49 minutes ago

  @Cloudflare you need to make D1 extremely reliable. Like a brick. At the edge. IDK how D1 just times out, errors randomly or apparently loses internet connection on a global anycast network? How does a SQL file do this? I don't get it. Errors listed below. CC @BraydenWilmoth

• 
  Boards (@Boards86) reported 57 minutes ago

  I gave her a Claude Pro referral link, and she built it all using their cloud hosted compute. Cloudflare free-tier, Github for versioning. Never once touched her laptop. Everything is a web app on her Pixel. She's officially Claude-pilled.

• 
  Ruben de Groot (@rubenpdegroot) reported 58 minutes ago

  @paoloanzn honestly t3code has been doing exactly this for me, i just got one 64gb ram vps on contabo and setup t3code there, i can just use their web ui from desktop or phone (behind cloudflare tunnel for security), you can just clone as many repos as you want ofc, and i just created a skill on how the t3api works such that claude code can just spin up threads inside of it no need for multiple VMs , just one big VPS only thing is that they dont support pi

• 
  Mark Lyck (@MarkLyck) reported 58 minutes ago

  @leerob @poteto @ericzakariasson can you please improve these simple Cursor PR features? Tried switching to Cursor this month, but there's a lot of missing / annoying stuff. 1. Add a preview link button. If the PR has a preview deployment from Vercel/Cloudflare/Netlify etc. Just give me a button I can click to quickly go to that preview deployment in Cursor or Browser. I don't want to have to go into the PR and go through review comments to find it. 2. Give me a keyboard shortcut to copy the Github PR link. Right now I have to click the chat, wait for it to load, then click "View PR" then click the 3 dots next to the github logo, then click copy link. That's way too many steps to share a Github/Origin link. 3. Why can't I configure trusted domains in the agent mode? Everytime I click a link I get this error "It's outside Cursor's trusted set. Only continue if you trust the source." And yes I did add it to my trusted domains in the IDE mode, but doesn't seem to work in agent mode. 4. Give me an indicator somewhere that shows how many approvals my PR has gotten if it has any. Right now the fastest way is to go to the github link, second fastest way is to go to the review section and scroll down and count manually if the PR doesn't have a lot of discussions. These are easy UX wins that would really help. 🙏🏻

• 
  Aarsh (@aarshps) reported 59 minutes ago

  @juiceboy_of_abj Airtel wrapping Cloudflare signup in a dead page is the kind of ISP fingerprint you only notice when a login form never appears.