NGINX is a common part of modern web infrastructure for many organizations and sits in front of websites, applications, APIs, and other services. However, a new group of vulnerabilities has recently put many NGINX installations under scrutiny.

The disclosure known as NGINX2Root covers five vulnerabilities that can cause memory corruption in NGINX. The most serious issue is CVE-2026-123007, which affects installations with HTTP/2 enabled. It lets an unauthenticated attacker execute code through the NGINX worker process.

The vulnerabilities were disclosed on November 18, 2026, and NGINX released fixes that day. Regular server maintenance and security checks can help businesses identify and address NGINX vulnerabilities before they become a bigger infrastructure risk, which is a key part of Server Management services.

What Is NGINX2Root?

NGINX2Root is the name researchers gave to a group of five NGINX vulnerabilities:

  • CVE-2026-123007
  • CVE-2026-123633
  • CVE-2026-121810
  • CVE-2026-123634
  • CVE-2026-113425

These vulnerabilities affect different NGINX features and configurations. While some involve HTTP/2 or HTTP/3, others need specific proxy or resolver configurations. Additionally, one affects NGINX when it works as a mail proxy.

CVE-2026-123007 has received the most attention as it has wider exposure. It affects NGINX versions from 1.9.5 through 1.31.8 when HTTP/2 is enabled. In other words, the vulnerable code has existed since HTTP/2 support was introduced in NGINX.

Our experts would like to point out that an affected server does not always give an attacker root access right away. The demonstrated code execution first happens with the privileges of the NGINX worker process. Researchers also demonstrated a separate Linux kernel vulnerability that could be chained for higher privileges. Furthermore, one report describes that kernel flaw as an unpatched SCTP issue.

Is NGINX2Root a Serious Concern?

NGINX receives requests from the public internet before passing them to backend services. A flaw in this layer can therefore create a direct security risk. The applications behind NGINX do not need to be vulnerable themselves.

Attackers can trigger CVE-2026-123007 through HTTP/2 when the service is reachable. Security researchers demonstrated remote code execution during testing of this flaw.

No exploitation in the wild had been reported as of November 19. CISA’s Known Exploited Vulnerabilities catalog also did not list these flaws on that date.

Rapid7 has published a technical analysis and a public exploit for CVE-2026-123007. The public exploit is hard to use, but exposed systems still need prompt patching.

Which NGINX Configurations Are Affected?

The risk depends on the NGINX version and the features enabled in its configuration.

CVE Affected configuration
CVE-2026-123007 HTTP/2 enabled
CVE-2026-123633 HTTP/2 enabled and requests proxied to a backend under specific worker connection conditions
CVE-2026-121810 HTTP/3 enabled and NGINX proxies requests to an upstream
CVE-2026-123634 Resolver and proxy settings combined with subrequests
CVE-2026-113425 NGINX used as a mail proxy with specific authentication settings

The affected version ranges also differ between the five vulnerabilities. So you have to check both the installed version and the server configuration. In other words, disabling one feature does not remove the entire NGINX2Root exposure.

Which NGINX Versions Are Vulnerable?

The affected versions depend on the individual CVE, as listed below.

CVE-2026-123007 affects NGINX Open Source versions 1.9.5 through 1.31.8 when HTTP/2 is enabled. CVE-2026-123633 affects versions from 1.19.7 through 1.31.8 under its required configuration. CVE-2026-121810 affects versions from 1.25.0 through 1.31.8 when HTTP/3 is enabled.

CVE-2026-123634 affects versions from 0.6.18 through 1.31.8 under its resolver and proxy conditions. CVE-2026-113425 affects versions from 0.5.33 through 1.31.8 when the mail proxy is configured. Unpatched NGINX Plus builds are also affected when HTTP/2 is enabled.

NGINX released the following fixed versions on November 18, 2026:

  • NGINX 1.30.7 for the stable branch
  • NGINX 1.31.9 for the mainline branch

If you are using packages supplied by Linux distributions, you should also check their vendor’s security advisory. Package maintainers can backport fixes without changing the upstream version number. So a simple version check may not show that a server is patched.

Ubuntu and OpenResty had no fix as of November 18. The retired ingress-nginx controller will not receive a fix and serves HTTP/2 by default. Patch availability may have changed since then, so confirm it in the advisory.

How Can You Protect Against NGINX2Root?

The best response is to update NGINX to a fixed release. You can also install the security update from your operating system vendor. A good patching process should include the following five steps:

1. Identify exposed NGINX servers

Create an inventory of all internet-facing NGINX installations in your environment. Check web servers, reverse proxies, application gateways, containers, and any other NGINX hosts.

2. Check the NGINX version

Confirm the installed package and compare it against the relevant security advisory. Do not rely only on

nginx -v
if your system backports patches.

3. Review the configuration

Check if HTTP/2, HTTP/3, the NGINX resolver, proxy features, or mail proxy functionality are enabled. Your configuration determines exposure to several of the NGINX2Root vulnerabilities.

4. Apply the available security update

Upgrade to NGINX 1.30.7 or 1.31.9 depending on your branch. Otherwise, apply the patched package from your operating system vendor. Restart NGINX as required so the updated worker processes are loaded.

5. Monitor the server after patching

Review monitoring data for unexpected process activity and other signs of compromise. Access logs may show little evidence of an attack through CVE-2026-123007. Patching closes the vulnerability, but it does not remove evidence of an earlier attack.

Can Disabling HTTP/2 Prevent NGINX2Root?

Disabling HTTP/2 can reduce exposure to both CVE-2026-123007 and CVE-2026-123633. It does not address the other three NGINX2Root vulnerabilities, which have separate triggers.

This makes disabling HTTP/2 a possible temporary mitigation when an immediate update is not possible. It should not be treated as a replacement for applying the vendor’s security fixes.

On NGINX 1.25.1 and later, add

http2 off;
to each affected server block. On older versions, remove
http2
from the
listen
directive instead.

HTTP/3 and other affected NGINX features also need to be assessed separately.

Does Cloudflare Protect Against NGINX2Root?

Cloudflare says its HTTP proxy can shield origins from certain HTTP/2 and HTTP/3 attack paths. Those connections end at Cloudflare’s edge instead of reaching the origin NGINX server directly. This only holds when the origin is reachable solely through Cloudflare’s proxy.

This protection does not cover every NGINX2Root scenario, and the level depends on enabled features. Organizations should therefore continue to patch their NGINX installations even when their websites use Cloudflare.

Conclusion

In brief, NGINX2Root highlights the risks that can arise when vulnerabilities affect software at the infrastructure layer. Identifying exposed NGINX servers, checking configurations, applying security updates, and monitoring for unusual activity can help reduce the risk. Regular vulnerability checks also make it easier to identify issues before they affect critical services.

Bobcares can help businesses manage the security and maintenance of their NGINX servers through Server Management services. Our experts can assess NGINX configurations, identify vulnerable software versions, apply security patches, and monitor server activity.