NGINX: two HTTP/2 flaws allow remote code execution, CERT-FR issues an alert

Top-Secured cyber watch: independent analysis based on the sources listed below. Top-Secured is not affiliated with the companies mentioned and will never ask you for your bank details or passwords.

F5 fixed CVE-2026-123007 and CVE-2026-123633 on 18 November, two memory bugs in the HTTP/2 module of NGINX Open Source and NGINX Plus that an unauthenticated attacker can turn into code execution. CERT-FR published an alert the same day, notes that a proof of concept is public and advises disabling HTTP/2 until servers are updated.

What happened

On 18 November 2026, F5 released two security advisories for NGINX, and the project shipped NGINX 1.31.9 and 1.30.7 the same day. Both main flaws sit in the HTTP/2 module. CVE-2026-123007 is an integer underflow in flow control that leads to a heap overflow; CVE-2026-123633 is a use-after-free triggered when NGINX proxies HTTP/2 requests to a backend while a worker runs short of worker_connections. F5 rates both 8.1 (CVSS 3.1) and 9.2 (CVSS 4.0): the usual result is a worker crash, but code execution is possible when ASLR is disabled or can be bypassed. CERT-FR issued alert CERTFR-2026-ALE-017 and says it is aware of a public proof of concept for CVE-2026-123007. According to Rapid7, its own exploit defeats ASLR and obtains a shell without authentication on a Docker image running 1.31.8. No attack in the wild had been reported on 18 November.

Who is affected

Any NGINX with HTTP/2 enabled: according to the CVE records, Open Source from 1.9.5 (CVE-2026-123007) and from 1.19.7 (CVE-2026-123633) up to 1.30.6 and 1.31.8, and NGINX Plus before 37.0.8.1 and 37.3.0.1. That includes reverse proxies, Kubernetes ingress controllers and container images that embed NGINX. The new releases also fix eight other bugs, including an HTTP/3 use-after-free and an overflow with try_files and alias.

What to do now

  1. Find every instance: nginx -v on servers, plus the images and ingress controllers you run.
  2. Upgrade to NGINX 1.30.7 or 1.31.9 (Plus: 37.0.8.1 or 37.3.0.1), or to your distribution’s fixed package, then run systemctl restart nginx: a reload does not replace the binary.
  3. If you cannot patch today, disable HTTP/2: remove the http2 parameter from listen lines and any http2 on; (grep -rn http2 /etc/nginx/), then nginx -t && systemctl reload nginx. Check with curl -so /dev/null -w '%{http_version}\n' https://your-site/: it must print 1.1.
  4. Make sure ASLR is on: cat /proc/sys/kernel/randomize_va_space should return 2.
  5. Watch for exited on signal 11 in error.log: repeated worker crashes may indicate attempts.

Our take

NGINX sits in front of almost everything, often in places nobody lists: appliances, images, ingress. With a working exploit already described by researchers, the gap between disclosure and attacks may be short. Knowing where each NGINX runs and being able to restart it the same day is the real protection.

Sources

How we choose, verify and write our analyses →

Is your own server exposed? Run a free check — public information only, no intrusion:

Check a server Check a website

Worried your servers are exposed? Our team audits, patches and monitors Linux and cloud infrastructure 24/7.

Talk to an expert Our managed server services →