
A vulnerability with a name that understates its severity landed on November 5, 2026: CVE-2026-123158, dubbed 'SSHole' by the security community, is an authentication bypass in OpenSSH that allows an unauthenticated attacker to log into passwordless accounts on Alpine Linux and any other musl-based Linux distribution with PAM disabled. The flaw was discovered by researcher Ozgun Kultekin and traces back to a code path introduced in a 2016 patch for a different OpenSSH vulnerability. Within hours of public disclosure, Previdian's threat intelligence team reported active exploitation attempts including a confirmed root compromise of a research honeypot. For any organisation running Alpine Linux, particularly in containerised environments, Kubernetes clusters, or lightweight cloud deployments, this is an emergency patch event.
The vulnerability exploits a subtle logic flaw in how the openssh server handles authentication on systems built on Alpine Linux, where PAM is typically disabled by default. In that setup, sshd checks passwords directly through musl's crypt rather than PAM, and an overly long password can return *, the same value often stored in the password field for key-only or passwordless accounts. Under normal circumstances, an account with no password set should be impossible to authenticate against over SSH. SSHole breaks this assumption.
The attack requires three conditions to be met: the attacker must have network access to the SSH port, they must know a valid account name on the target system, and they must supply a password between 257 and 1,024 bytes in length. Attackers can exploit accounts with * as password hash only if at least one account on the same system has a real password set. When all these conditions align, OpenSSH incorrectly processes the authentication result and grants access to the passwordless account bypassing the fundamental guarantee that a locked account cannot be logged into.
The root cause traces to a 2016 patch for CVE-2016-6210, which introduced a code path in the musl-specific authentication handling that was never re-audited for the interaction with passwordless accounts. This is a reminder that security patches do not exist in isolation changes made to fix one vulnerability can introduce subtle weaknesses in adjacent code paths that go undetected for years.
OpenSSH released version 10.6p2 on November 5 the same day as disclosure which closes the bug. Alpine 3.16 through 3.24 are vulnerable to exploitation if misconfigured, and edge shipped the openssh server package as 10.6_p2-r0. Systems using glibc distributions and systems with PAM enabled, including openssh server pam setups, are not affected.
Alpine Linux may not be the first distribution that comes to mind when enterprise security teams enumerate their Linux estate, but its footprint in modern infrastructure is enormous. Alpine is the base image of choice for Docker containers across the global DevOps ecosystem its minimal size (under 5MB) and security-focused design made it the standard for Kubernetes pods, serverless functions, CI/CD pipelines, and containerised microservices, and some official Alpine cloud images also deserve review when SSH is enabled, while systems provisioned with cloud-init are not affected in the same way. Many organisations running Alpine have no direct inventory of which containers use it as a base image.
This creates a hidden attack surface problem. An enterprise may have patched every visible Alpine Linux server but still be running hundreds or thousands of Alpine-based container images that have not been rebuilt since the vulnerability was disclosed. Container images are not automatically updated when the underlying package is patched, they must be explicitly rebuilt from updated base images and redeployed; a running container upgraded in place will lose that change on restart, so teams should rebuild from a fixed base image rather than rely on ad hoc patching, especially because older Alpine-based images may have no fixed package available. Organisations that rely on automated container scanning tools need to confirm those tools are flagging the OpenSSH version in every running image, not just the host OS.
The SSH daemon is not always exposed externally in containerised environments, which reduces but does not eliminate the attack surface. Within a Kubernetes cluster, an attacker who has gained a foothold in one pod and can reach the SSH port of another pod over the internal network has a viable lateral movement path using SSHole against any Alpine-based container running an SSH daemon with the vulnerable configuration. Only workloads that run sshd are exposed; images that do not start the service are not affected.

Previdian's threat intelligence report on November 5 documented exploitation attempts against instrumented SSH honeypots within hours of disclosure. One honeypot running a vulnerable Alpine Linux configuration with an intentionally exposed SSH port recorded a successful authentication bypass that resulted in the attacker achieving root-level access. The speed from disclosure to confirmed exploitation measured in hours, not days sets SSHole apart from many vulnerabilities that take weeks to see meaningful attack activity.
The attack pattern observed in the honeypot data follows a recognisable playbook: automated scanning tools enumerate SSH-exposed hosts, probe for vulnerable OpenSSH versions, attempt the authentication bypass with crafted long passwords, and upon success, deploy a dropper that establishes persistence and beacons back to attacker infrastructure. This is not a sophisticated manual attack, it is a scripted, industrialised operation that scales to hundreds of thousands of targets simultaneously.
For SOC teams, the honeypot data provides actionable detection intelligence. SSH authentication attempts using very long password strings (257+ bytes) against accounts that should not be authenticatable are a strong signal of SSHole exploitation attempts. Analysts should also search SSH daemon logs during the exposure window for Accepted password for entries, especially successful password logins for accounts that should not allow password logins; checking /etc/shadow can confirm whether an affected account had * as its stored password. Failed attempts are valuable too, they indicate your environment is being scanned even if the attack fails or has not yet succeeded.
Immediate patch deployment is the first priority. Update OpenSSH to 10.6p2 on all Alpine Linux hosts, and rebuild and redeploy all Docker and container images based on Alpine to pick up the patched OpenSSH package. After upgrading the package, restart sshd so the running daemon stops using the old code. If patching is delayed, set PasswordAuthentication to no as a temporary mitigation. Confirm that your container registry is scanning for the vulnerable version and blocking deployment of unpatched images.
For detection, SHELT's SOC has deployed rules covering three key signals: SSH authentication attempts with anomalously long password strings, authentication events against system accounts or service accounts that should never authenticate interactively, and post-exploitation indicators including unexpected process spawns from the SSH daemon, new cron entries, or outbound connections from containers that do not normally make external network calls.
Threat hunting in environments where Alpine is widely used should focus on sshd logs for the period between November 5 and your patch deployment date. Any successful authentication event during that window that cannot be attributed to a known user action should be treated as a potential compromise and investigated. SHELT's SOC team can conduct a targeted threat hunt across your environment to establish a clean bill of health before the exploitation window widens further.
SHELT's Security Operations Center operates a continuous vulnerability intelligence programme that moves from public disclosure to client advisory and detection rule deployment within the same operational cycle. When SSHole was disclosed on November 5, SHELT's threat intelligence team immediately assessed client exposure across all monitored environments, identified Alpine Linux deployments and container images in scope, and pushed updated detection rules covering the specific SSH authentication patterns associated with exploitation attempts.
For clients who cannot immediately patch due to change management constraints, containerised environment complexity, or operational dependencies SHELT implements network-level compensating controls: restricting SSH access to known management source IPs, disabling password-based SSH authentication in favour of key-only authentication (which is not affected by this vulnerability), and increasing log collection and alerting sensitivity for SSH events on affected hosts.
SHELT's SOC service provides this rapid-response capability to organisations across MENA and globally that lack the in-house expertise to triage new critical vulnerabilities at the speed that modern threat actors exploit them. The window between disclosure and active exploitation is measured in hours for vulnerabilities of this severity and your SOC response needs to match that pace.
No. CVE-2026-123158 affects OpenSSH only on musl libc-based systems with PAM disabled. Ubuntu, Debian, Red Hat, CentOS, and Amazon Linux are glibc distributions with PAM enabled by default, unlike systems built on Alpine Linux and other musl-based targets. They are not affected, and non Linux systems are also not affected. The primary impact is on Alpine Linux hosts and Alpine-based Docker containers.
Partially. If the SSH daemon is not externally exposed, remote exploitation from the internet is not directly possible. However, an attacker with any foothold inside your network or Kubernetes cluster can potentially use SSHole for lateral movement against Alpine-based containers that run sshd. Patching all container images remains the correct response regardless of external exposure.
Container scanning tools such as Trivy, Snyk Container, and AWS ECR image scanning can identify the OpenSSH version in running and stored images. Run a scan immediately targeting your container registry and running pods. Patching a running container is only temporary, so rebuilding from a fixed base image is preferred. Any image with the vulnerable openssh server package on Alpine 3.16 through 3.24 is in scope; images on older Alpine releases have no fixed package, so rebuild from the updated Alpine base image and redeploy.
Set PasswordAuthentication no to disable password logins and switch to key-only authentication, as SSHole requires password-based authentication and does not affect key-based login. A second command should also verify the final effective SSH configuration, because Match blocks can turn password logins back on for specific users or source addresses. PermitRootLogin prohibit-password does not allow root password login; the bigger risk is other accounts that still permit password logins. Additionally, restrict SSH access to known source IPs via firewall rules. These are compensating controls, not a substitute for patching apply them while your patching window is arranged.
Yes. SHELT offers a targeted threat hunt for SSHole post-exploitation indicators, covering SSH authentication logs, including searches for Accepted password for entries in sshd logs correlated with accounts that show * in /etc/shadow, container process trees, cron and systemd persistence mechanisms, and outbound network connections from affected hosts. Special review is also warranted for Alpine cloud images using the default alpine user, because that account may have passwordless doas; if compromise is confirmed, verify whether an attacker could run doas to reach root. Contact SHELT to initiate a rapid assessment if you had Alpine Linux systems exposed between November 5 and your patch date.
.png)
© SHELT 2023 Privacy Policy | Terms & Conditions