Critical NGINX HTTP/2 Vulnerability CVE-2026-123007 Allows Remote Code Execution and Potential Root Access

A significant security crisis has emerged for the global web infrastructure as F5, the developer of the ubiquitous NGINX web server, has disclosed a series of high-severity vulnerabilities affecting its core software. At the center of this discovery is CVE-2026-123007, a critical flaw within NGINX’s HTTP/2 implementation that enables unauthenticated attackers to achieve remote code execution (RCE). This vulnerability, colloquially dubbed "nginx2root," has the potential to impact a vast majority of the internet, given that NGINX powers a significant percentage of the world’s websites, load balancers, and reverse proxies.
The vulnerability stems from a decade-old oversight in the handling of HTTP/2 traffic, a protocol designed to optimize web performance. Since its introduction in version 1.9.5 on September 22, 2015, the flawed code has remained embedded in the open-source distribution. The discovery of this vulnerability serves as a stark reminder of the risks associated with legacy code in high-performance networking software.
A Chronology of the Vulnerability and Disclosure
The existence of CVE-2026-123007 was confirmed following coordinated research efforts by security firms Wiz and Xint. The timeline of the discovery and subsequent patching indicates a rapid response from F5, though the sheer scale of the deployment makes remediation a complex challenge.
The vulnerability was present in all open-source releases from version 1.9.5 through 1.31.8. On October 14, 2026, F5 issued security releases 1.31.7 and 1.30.6 to address earlier concerns, but the critical "nginx2root" flaw was only fully rectified in the most recent updates—versions 1.31.9 and 1.30.7—released today.
Security researchers at Wiz successfully demonstrated that they could bypass Address Space Layout Randomization (ASLR)—a common memory protection technique—to execute arbitrary code. The Xint research team further corroborated this by executing code with nothing more than a reachable HTTP/2 port, highlighting the ease with which a sophisticated actor could theoretically compromise a server.
Technical Mechanics: How the Attack Works
The "nginx2root" vulnerability functions by exploiting the way NGINX processes HTTP/2 frames. When an attacker sends specifically crafted malicious requests to an NGINX server with HTTP/2 enabled, they can trigger an integer underflow condition. This manipulation of memory allows the attacker to execute code within the context of the NGINX worker process.

While executing code as a worker process is dangerous, the severity increases exponentially when combined with other system-level vulnerabilities. The Computer Incident Response Center Luxembourg (CIRCL) has warned that by chaining this exploit with a separate bug in the Linux kernel’s Stream Control Transmission Protocol (SCTP) implementation, an attacker can escalate their privileges from an NGINX worker to root access on the host machine. This remains a threat even when NGINX is isolated within a Docker container, demonstrating the danger of "container breakout" scenarios.
The attack is not trivial to execute, requiring specific hardware conditions. Rapid7’s technical analysis suggests that the exploit requires approximately 128 GB of HTTP/2 traffic to be successful, a process that takes roughly two and a half minutes. Furthermore, the exploit is highly dependent on specific processor instructions, namely avx512f and avx512vl, found on certain modern x86-64 CPUs. Despite these requirements, the public availability of proof-of-concept (PoC) code on GitHub has significantly lowered the barrier to entry for malicious actors.
A Multitude of Flaws
While CVE-2026-123007 has captured headlines, F5 has addressed a total of five distinct vulnerabilities in the latest patch set. Each of these carries a CVSS 4.0 score of 9.2, signaling a "critical" severity rating. The landscape of these flaws is summarized as follows:
- CVE-2026-123007: The primary HTTP/2 RCE flaw.
- CVE-2026-123633: Affects configurations where HTTP/2 is enabled and requests are proxied to a backend server while the worker is experiencing high connection pressure (low
worker_connections). - CVE-2026-123634: A vulnerability triggered when a proxied server name is dynamically built from variables and combined with Server-Side Includes (SSI) or mirror subrequests.
- CVE-2026-121810: A flaw specifically affecting implementations where HTTP/3 is enabled.
- CVE-2026-113425: A vulnerability occurring when NGINX is utilized as a mail proxy.
The breadth of these vulnerabilities suggests a systemic issue in how NGINX handles various protocols and dynamic configuration parameters, necessitating a comprehensive audit for any organization relying on NGINX as their primary web gateway.
Mitigation and Remediation Strategies
F5 has been explicit in its guidance: organizations must upgrade their NGINX instances immediately. For NGINX Open Source, users should move to 1.31.9 or 1.30.7. For NGINX Plus users, the patches are integrated into releases 37.3.0.1 and 37.0.8.1.
Crucially, administrators are warned that a standard "reload" command is insufficient. Because a reload does not stop the existing worker processes, the vulnerable code would remain active in memory. A full restart or a live binary upgrade is required to ensure the old, vulnerable processes are terminated and replaced by the patched version.
For those unable to patch immediately, temporary workarounds exist, though they come with performance trade-offs. The most effective mitigation is the complete disabling of HTTP/2 and HTTP/3 support across all server blocks. This can be achieved by removing the http2 directive from listen statements and ensuring http2 off; is set. While this will impact the performance benefits provided by these protocols, it effectively neutralizes the primary vectors for the most critical flaws.

For the Linux kernel SCTP vulnerability, CIRCL recommends blocking the SCTP kernel module entirely by adding install sctp /bin/false to a configuration file within /etc/modprobe.d/. This serves as a vital "defense-in-depth" measure, preventing the secondary privilege escalation required to move from an NGINX worker to root host access.
Broader Implications for Web Infrastructure
The discovery of "nginx2root" highlights the fragility of the modern web stack. NGINX acts as the gatekeeper for millions of applications; its compromise is essentially a "master key" event for any infrastructure relying on its default configurations.
The fact that this flaw remained hidden for a decade in a foundational piece of open-source software raises broader questions regarding the security audit practices for critical internet infrastructure. As the industry shifts toward more complex protocols like HTTP/3 and increasingly relies on containerized microservices, the attack surface for web servers continues to expand.
Organizations are encouraged to monitor their infrastructure for anomalous behavior. CIRCL has suggested several indicators of compromise (IoCs), including:
- NGINX worker processes initiating unexpected sub-processes.
- Frequent abnormal exits or unexpected core dumps from worker processes.
- Persistent, long-lived HTTP/2 connections that exhibit heavy inbound traffic but minimal logged request activity—a sign that an attacker may be using control frames to hide their movements.
- Unusual outbound network activity or DNS queries originating from the web server.
As of the current reporting period, there have been no confirmed reports of active exploitation in the wild. However, the publication of PoC code means that the "window of opportunity" for attackers is closing rapidly. System administrators and DevOps engineers are advised to prioritize these updates in their current sprint cycles to avoid becoming the next victim of a mass-exploitation campaign. The incident serves as a definitive case study in the necessity of robust patch management and the risks inherent in relying on legacy protocol implementations without continuous security scrutiny.







