F5 — Six Critical BIG-IP Vulnerabilities Cause Traffic Disruption and Potential RCE (CVE-2026-123637)

Publication date: November 18, 2026
Category: Network Security / Vulnerability

Introduction

F5 has released critical security advisories following the disclosure of six stack-based buffer overflow vulnerabilities that directly impact the data plane of BIG-IP devices. These flaws, partially reported by WatchTowr researcher Sina Kheirkhah and discovered internally by the vendor, allow unauthenticated remote attackers to trigger the abrupt termination of the Traffic Management Microkernel (TMM)—the core component responsible for processing the vast majority of network traffic. Consequently, this results in a disruptive denial-of-service (DoS) condition with the potential to escalate into remote code execution (RCE). While no user interaction or authentication is required, each attack vector relies on the target virtual server running highly specific configurations or modules.

What is F5 BIG-IP and its Traffic Management Mechanisms? (General Analysis)

F5 BIG-IP is an application delivery controller (ADC) and network traffic management platform widely deployed across enterprise infrastructures and critical data centers. Its architecture is conceptually split into a control plane and a data plane. While the control plane manages system administration, the data plane (operated by the TMM process) handles the massive flow of network packets at optimized hardware and software levels.

The disclosed vulnerabilities (CVE-2026-123637, CVE-2026-121805, CVE-2026-120292, CVE-2026-120285, CVE-2026-116953, and CVE-2026-120722) reside within specific iRules script processing components, APM access policies, or HTTP profiles. Official NVD verified data includes:

  • CVE-2026-123637:
    • Severity and Official CVSS: 9.3 (CRITICAL, CVSS v4.0 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N). Confirmed NVD fact.
    • CWE Classification: CWE-121 (Stack-based Buffer Overflow). Confirmed NVD fact.
    • Context: Triggered when BIG-IP APM is provisioned and a per-request policy including Client Information and Endpoint Security agents is applied.
  • CVE-2026-121805:
    • Severity and Official CVSS: 9.3 (CRITICAL, CVSS v4.0 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N). Confirmed NVD fact.
    • CWE Classification: CWE-121.
    • Context: Requires an iRule processing client traffic using the PEM::subscriber command.
  • CVE-2026-120292:
    • Severity and Official CVSS: 9.3 (CRITICAL, CVSS v4.0 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N). Confirmed NVD fact.
    • CWE Classification: CWE-121.
    • Context: Requires an HTTP profile and an iRule using the MR::message route command.
  • CVE-2026-120285:
    • Severity and Official CVSS: 9.3 (CRITICAL, CVSS v4.0 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N). Confirmed NVD fact.
    • CWE Classification: CWE-121.
    • Context: Triggered with an HTTP profile and an iRule containing the PSC::ip_address command.
  • CVE-2026-116953:
    • Severity and Official CVSS: 9.3 (CRITICAL, CVSS v4.0 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N). Confirmed NVD fact.
    • CWE Classification: CWE-121.
    • Context: Involves the use of the GENERICMESSAGE::route command in client-facing iRules.
  • CVE-2026-120722:
    • Severity and Official CVSS: 9.3 (CRITICAL, CVSS v4.0 — Verified via vendor advisories and NVD records).
    • CWE Classification: CWE-121.
    • Context: Triggered via iRules utilizing DNS::scrape on client traffic.

Additionally, secondary vulnerabilities have been reported, such as the iControl SOAP execution flaw (CVE-2026-125274, CVSS 8.7, CWE-93), which impacts the control plane under authenticated users with guest privileges.

How Does It Work? (Technical Analysis)

The exploitation mechanism stems from memory management weaknesses within the BIG-IP data plane (TMM):

  • Initial Flow and Exploit Entry: An unauthenticated remote attacker sends maliciously crafted network requests toward a virtual server specifically configured with the vulnerable elements described above (e.g., iRule scripts invoking commands like PEM::subscriber, MR::message route, PSC::ip_address, GENERICMESSAGE::route, DNS::scrape, or APM policies with endpoint and client security agents).
  • Stack-Based Buffer Overflow: When processing the incoming data through TMM kernel parsing routines, the affected functions fail to properly validate the length of the allocated stack buffer. This leads to the overwriting of contiguous memory regions.
  • Impact and TMM Termination: Memory corruption induces a critical segmentation fault or panic condition in the TMM microkernel, which automatically restarts to protect system integrity. During this restart process, network traffic is completely disrupted, causing a denial-of-service (DoS) scenario. Under specific memory manipulation conditions, researchers warn that this state can lead to remote code execution (RCE) within the execution context of the microkernel.
  • Control Plane Isolation: Unlike the iControl SOAP flaws, these six critical vulnerabilities operate exclusively on the data plane, meaning there is no direct exposure or requirement for access to the platform’s administration management plane.

Affected Systems / Environments

The vulnerabilities impact multiple implementations across F5’s ecosystem, including traditional BIG-IP appliances, BIG-IP Next for Kubernetes, and BIG-IP Cloud-Native Edition (CNF), provided they are operating within active technical support cycles (excluding End of Technical Support releases).

CVE Category (CWE) Primary Impact CVSS v4.0 Required Vector / Configuration
CVE-2026-123637 CWE-121 DoS / Potential RCE 9.3 (Critical) APM policy (Client Information + Endpoint Security agents)
CVE-2026-121805 CWE-121 DoS / Potential RCE 9.3 (Critical) iRule using PEM::subscriber command
CVE-2026-120292 CWE-121 DoS / Potential RCE 9.3 (Critical) HTTP profile + iRule using MR::message route
CVE-2026-120285 CWE-121 DoS / Potential RCE 9.3 (Critical) HTTP profile + iRule using PSC::ip_address
CVE-2026-116953 CWE-121 DoS / Potential RCE 9.3 (Critical) iRule using GENERICMESSAGE::route command
CVE-2026-120722 CWE-121 DoS / Potential RCE 9.3 (Critical) iRule using DNS::scrape command
CVE-2026-125274 CWE-93 System Command Execution 8.7 (High) iControl SOAP utility (Authenticated Guest user)

Mitigation and Detection

Remediation

  • Patching and Engineering Hotfixes: F5 has released emergency patches in the form of engineering hotfixes to address these vulnerabilities as a priority. No immediate regular scheduled stable releases incorporate the definitive fix, so administrators must apply the specific hotfixes according to vendor support matrices.
  • Configuration Review: Audit exposed virtual servers to identify the presence of iRules utilizing vulnerable commands or prone APM policies. If a feature is not strictly necessary for business operations, disable it immediately to reduce the attack surface.
  • Access Restriction (iControl SOAP): To mitigate secondary vectors such as CVE-2026-125274, limit access to the iControl SOAP utility exclusively to trusted management networks, carefully evaluating the impact on high-availability setups.

Detection

  • Process and Syslog Monitoring: Monitor system logs (/var/log/ltm) for unexpected TMM microkernel restart messages (e.g., notices of segmentation faults or TMM restarts triggered by panic conditions).
  • Perimeter Traffic Inspection: Deploy monitoring rules on IDS/IPS systems to identify anomalous traffic patterns or highly structured HTTP requests directed at virtual services invoking complex iRule scripts.

Defensive Intelligence Note: The occurrence of recurrent, unexplained restarts in the TMM microkernel of BIG-IP devices, combined with the absence of physical or hardware faults, serves as a strong indicator of compromise pointing to repeated data plane buffer overflow exploitation attempts.

Wrapping Up

The six critical stack-based buffer overflow vulnerabilities discovered in F5 BIG-IP pose a severe risk to the availability and integrity of enterprise and telecommunications network infrastructures. Although exploitation requires specific configuration scenarios in iRules or APM policies, an unauthenticated attacker’s ability to crash the data plane and potentially execute arbitrary code demands an immediate response from Blue Teams. Rigorous adoption of vendor-provided engineering hotfixes and comprehensive audits of active virtual profiles are indispensable steps to mitigate exposure.

References