Skip to main content

F5’s Six Critical BIG-IP Flaws Have Hotfixes Only, and Whether You Are Exposed Depends on Your iRules

F5 disclosed six critical BIG-IP buffer overflows on 18 November and fixes only through engineering hotfixes. Each needs a specific iRule or APM policy, so the first job is finding out whether you have one.

KBY News Desk7 min read
Close-up of a TP-Link Ethernet switch with yellow, red, and white cables connected.

On Wednesday 18 November F5 disclosed six critical flaws in the data plane of BIG-IP, the load balancer and application delivery platform that sits in front of many large organisations’ websites and VPNs, and as of Saturday it had no scheduled software release that fixes them. All six are stack-based buffer overflows. F5 says an unauthenticated attacker can crash the Traffic Management Microkernel, known as TMM, which is the part of BIG-IP that handles traffic, and that the flaws “may also allow remote code execution”. F5 rates each 9.8 on the CVSS v3.1 scale and 9.3 on v4.0. Neither F5 nor anyone else reports exploitation so far, and none is on CISA’s Known Exploited Vulnerabilities list.

The detail that decides who needs to hurry is that none of the six is reachable on every BIG-IP. Each needs a specific configuration to be in place, mostly an iRule (a small script that tells BIG-IP how to treat traffic) that calls one particular command. That makes this a story about finding out whether you are exposed before you decide how fast to move.

#What was disclosed, and what each flaw needs

The Hacker News, which went through the advisories on the day, lists the six as follows. We have not been able to read F5’s own advisory pages, which load only in a browser, so every detail below is attributed to that report and checked against the H-ISAC bulletin and a second security outlet, Daily Security Review, which agree on the scores, the count and the lack of exploitation. Read F5’s notification, which The Hacker News links as K000164058, before you act on any version number.

  • CVE-2026-123637 affects BIG-IP Access Policy Manager (APM) when a per-request policy uses the Client Information and Endpoint Security agents.
  • CVE-2026-121805 needs an iRule that uses the PEM::subscriber command on client traffic. It affects BIG-IP Policy Enforcement Manager and BIG-IP Next for Kubernetes, and, according to The Hacker News, BIG-IP CNF.
  • CVE-2026-120292 needs an HTTP profile and an iRule using MR::message route, and affects all BIG-IP modules and BIG-IP Next for Kubernetes.
  • CVE-2026-120285 needs an HTTP profile and an iRule using PSC::ip_address, and affects all BIG-IP modules, BIG-IP Next for Kubernetes and CNF.
  • CVE-2026-116953 needs an iRule using GENERICMESSAGE::route on client traffic, with the same product list as CVE-2026-120292.
  • CVE-2026-120722 needs an iRule using DNS::scrape on client traffic and affects BIG-IP DNS and BIG-IP Next for Kubernetes. The Hacker News notes that F5’s advisory lists CNF as not vulnerable while the CVE record, updated on 18 November, lists it as affected. Treat CNF as in scope until F5 resolves that.

Five of the six therefore come down to an iRule that uses one named command. F5 says the management interface is not exposed, so the risk sits on the traffic side: a virtual server that accepts connections and runs one of those iRules.

#The fix is a hotfix, and a hotfix is a decision

The awkward part is how F5 is fixing them. According to The Hacker News, F5 offers engineering hotfixes only, with no workaround and no date for a scheduled release that contains the fixes. The report names hotfixes for BIG-IP Next for Kubernetes (BNK-2.4.2-EHF1, BNK-2.3.4-EHF1 and BNK-2.2.5-EHF1) and for BIG-IP CNF (CNF-2.4.2-EHF1, CNF-2.3.4-EHF1 and CNF-2.2.5-EHF1, for the two CVEs that touch CNF). It does not list build numbers for classic BIG-IP in the part we could read, so find yours in F5’s advisory.

F5’s own caution, quoted by The Hacker News, is that “engineering hotfixes do not undergo the extensive QA assessment of scheduled software releases”. The H-ISAC bulletin makes the same point. Hotfixes carry no warranty, F5 checks only supported versions, and a system already running a custom hotfix needs a new one from F5 support rather than a simple install on top.

That turns a patch job into a risk trade-off. We think the sensible reading is that there are two groups. If a BIG-IP faces the internet and runs one of the five iRule commands, or the APM policy in the first flaw, the exposure is real and the unauthenticated crash is the least that can happen, so get the hotfix tested and in place quickly even though it has had less testing. If the matching configuration is on a BIG-IP that only internal users can reach, or it is not there at all, waiting for F5 to put the fix into a scheduled release is a defensible choice, provided you write the decision down and watch for exploitation reports.

#How to find out whether you are exposed

The configuration triggers are specific enough to search for. This is our suggested order, not F5’s:

  1. List your iRules and search them. Export or list the iRules on each BIG-IP and search for the five commands: PEM::subscriber, MR::message route, PSC::ip_address, GENERICMESSAGE::route and DNS::scrape. A match tells you which virtual servers to look at next, because an iRule only matters if it is attached to one that carries client traffic.
  2. Check which of those virtual servers face the internet. Those go to the front of the queue.
  3. Check APM per-request policies. Look for any policy that uses both the Client Information and Endpoint Security agents. If you run APM for remote access, you are the most likely to have one.
  4. Check the platform. BIG-IP Next for Kubernetes and CNF are on the affected list, so a team that runs either needs its own hotfix from the list above, and should ask F5 about the CNF discrepancy.
  5. Ask F5 support for the hotfix for your version and ask them to confirm whether any custom hotfix you already run is replaced.
  6. Watch for crashes. A TMM restart in the logs on a virtual server that matches the configuration is worth an investigation now, not after the patch.

If your estate is mainly NGINX rather than BIG-IP, note that F5 published 22 advisories on 18 November and ten of them concern NGINX, according to The Hacker News and Daily Security Review. Our earlier piece on the NGINX HTTP/2 flaws covers that product.

#Why this one is not yet an emergency, and what would change our mind

Nobody has reported attacks, there is no public exploit in the reporting we read, and each flaw needs a configuration that many deployments will not have. Against that, we think the pattern matters. In September an APM flaw, CVE-2026-94127, was exploited and added to CISA’s catalogue, and the same reports are careful to separate it from these six. A crash bug that an unauthenticated attacker can trigger over the network is the kind of thing that attracts attention quickly, and four of the six were reported by an outside researcher, Sina Kheirkhah of watchTowr, so the detail is already in more than one set of hands. We read that as a reason to expect public technical write-ups, though that is our judgement, not a report.

We would change our assessment if any of three things happened: F5 or CISA reports exploitation, a public proof of concept appears, or F5 confirms that remote code execution is achievable in practice rather than possible in principle. Any of those moves a matching internet-facing BIG-IP from “test the hotfix this week” to “fix tonight”.

#What this means for smaller organisations

BIG-IP is enterprise kit, and most small businesses will never own one. If your IT is outsourced, the useful question to your provider is whether any client sits behind a BIG-IP they manage, and when they will tell you what they found. Anyone who has an application hosted by a larger partner can ask the same of the partner. There is nothing for a firm with no F5 equipment to do.

#The question to put to F5 or your reseller

Ask, in these words: “Which of our BIG-IP systems run an iRule or APM policy matching the six advisories, which hotfix build do we need for each version, and what is your guidance on running an engineering hotfix on a production system until a scheduled release exists?” If the answer is vague, that is information. For the wider routine around triage and tracking, our entry on vulnerability management sets out the vocabulary.

Evidence trail

Sources and verification

Primary documentation and external technical references used in this article.

  1. 01The Hacker News: F5 Warns 6 Critical BIG-IP Flaws Can Crash Traffic in Specific Setups, May Allow RCE (18 November 2026)thehackernews.com
  2. 02H-ISAC Vulnerability Bulletin: Six Critical F5 Big-IP Flaws Could Lead to RCEaha.org
  3. 03Daily Security Review: F5 Discloses Six Critical BIG-IP Buffer Overflows That Crash TMM (19 November 2026)dailysecurityreview.com

About this page

Credited to

Dates

Published

Corrections

No corrections are listed for this page. Some review sweeps corrected pages without naming each one; those are in the corrections log.

Reader Interaction

Comments

Add a thoughtful note on F5’s Six Critical BIG-IP Flaws Have Hotfixes Only, and Whether You Are Exposed Depends on Your iRules. Comments are checked for spam and held for moderation before appearing.

Loading comments...

Get new fixes by email.

New runbooks, configuration traps and fixes from KBY Technologies, about twice a week.