BREAKING: Cloudflare MINOR — HTTP/3 Issue Affecting R2 Custom Domains [8x66bpk6p9kk]

Cloudflare is experiencing a partial outage: HTTP/3 disruption impacting R2 custom domains. Immediate workarounds for indie hackers inside.

BREAKING: Cloudflare HTTP/3 Issue Affecting R2 Custom Domains

Status: INVESTIGATING | Severity: MINOR | Impact: PARTIAL

Cloudflare is currently experiencing an HTTP/3 connectivity issue specifically affecting R2 (object storage) requests through custom domains. This is NOT affecting all Cloudflare services—most traffic is unimpacted. However, if you're serving assets via R2 custom domains, you need to act now.

---

What's Down & Who's Affected

Affected:

HTTP/3 requests to R2 custom domains (e.g., assets.yourcompany.com → R2 bucket)

Clients forcing HTTP/3 via QUIC protocol

Static asset delivery, CDN-cached content, and media files through R2 endpoints

NOT Affected:

HTTP/2 or HTTP/1.1 traffic (fallback works)

Direct R2 API calls

Standard Cloudflare cache/CDN

Workers, Pages, or other Cloudflare products

---

Immediate Workarounds (DO THIS NOW)

1. Force HTTP/2 Fallback

Add to your Cloudflare DNS settings for R2 custom domains:

Disable HTTP/3 in Cloudflare dashboard → Networks tab → Toggle off "HTTP/3 (QUIC)"

Clients will automatically fall back to HTTP/2 (near-zero latency impact)

2. Use Direct R2 Endpoint (Temporary)

Switch asset URLs from custom domain to direct R2 endpoint: ``` // Change from: https://assets.yoursite.com/image.jpg

// Temporarily use: https://mybucket.eu.r2.cloudflarely.com/image.jpg ``` This bypasses the custom domain routing entirely.

3. Cache Aggressively

Increase Cloudflare cache TTL temporarily: ``` Cache-Control: public, max-age=86400 ``` This reduces origin requests during the incident.

---

How to Check If You're Affected

Run this diagnostic: ```bash

Test HTTP/3 connectivity

curl -I --http3 https://assets.yoursite.com/image.jpg

Test HTTP/2 fallback

curl -I --http2 https://assets.yoursite.com/image.jpg

Check if HTTP/3 is enabled

grep -i "alt-svc" <(curl -I https://assets.yoursite.com/image.jpg) ```

If HTTP/3 fails but HTTP/2 succeeds → you're affected.

Check your logs:

Cloudflare Analytics → Requests with protocol: QUIC showing errors

Application logs for increased latency on asset delivery

---

Alternative Tools to Consider

If you want redundancy during incidents:

AWS CloudFront + S3: Native HTTP/3, battle-tested CDN

Bunny CDN: Cheaper R2 alternative, independent HTTP/3 stack

AWS S3 + Lambda@Edge: More control, higher complexity

Fastly: Enterprise-grade, excellent HTTP/3 support

*Note: Most indie hackers should stick with Cloudflare R2 once resolved—this is a rare incident.*

---

Monitor for Recovery

Check status here:

[Cloudflare Status Page](https://www.cloudflarestatus.com)

Look for "R2" or "HTTP/3" incident updates

Follow [@Cloudflare](https://twitter.com/Cloudflare) for updates

Estimated resolution: Cloudflare typically resolves partial issues within 2-4 hours. Full investigation may take 12-24 hours.

What to do after resolution: 1. Re-enable HTTP/3 in Cloudflare dashboard 2. Monitor asset delivery for 30 minutes 3. Revert to custom domain URLs 4. Document incident for your runbook

---

Bottom Line

This is fixable in 2 minutes. Disable HTTP/3 or use direct R2 URLs. Your site won't go down. Most users won't notice because HTTP/2 is nearly identical in performance. Stay calm, implement one workaround, and move on.

We'll update this as Cloudflare provides more details.