← All updates

Security / News explained

OpenSSH on Alpine: the login path determines exposure

CISA added the OpenSSH password bypass to its exploited-vulnerability catalog on November 9. Alpine’s advisory identifies specific login conditions, branch-specific fixes and a restart requirement that matters when closing the change.

By Zyel · Published · 4 min read

CISA added OpenSSH vulnerability CVE-2026-123158 to its Known Exploited Vulnerabilities catalog on November 9, 2026. The important question for operators is which login path a particular server uses, followed by whether the running service contains the fix. [3]

Portable OpenSSH 10.6p2, released November 5, includes a critical workaround for a password-authentication bypass affecting some Linux systems using musl libc with PAM disabled. The OpenSSH team places the root cause outside sshd, while explaining that sshd can avoid the exploitable behavior. [1]

This is a configuration-dependent problem. Treating every machine with an OpenSSH package as equally exposed would obscure the investigation. Treating a familiar key-based login as proof of safety would also skip the question that matters: what other authentication routes remain available on that server?

The account configuration matters

Alpine’s November 5 advisory identifies a combination: the non-PAM server path, password authentication enabled, an account whose password field is an asterisk, and another local account with a supported SHA-256 or SHA-512 password hash. A PAM-enabled server running with UsePAM yes is outside that affected configuration. [2]

The advisory notes that key-only accounts can meet the account condition. OpenSSH’s default root-login restriction protects root from this password route, but does not establish protection for other eligible accounts. The flaw is not described as a way to break an SSH key. [2]

Imagine a hypothetical operations team that always signs in with keys and describes its environment as key-only. That phrase might describe staff habits. A useful exposure decision needs to describe the server’s effective policy as well. The distinction is between an access method people choose and the methods the service is willing to accept.

Our interpretation is that inventory needs enough detail to answer that policy question. A product name and a port number offer a starting point. The package build, operating-system branch and effective authentication settings determine whether that starting point represents the affected combination. A clear inventory helps an operator explain why two superficially similar machines receive different decisions.

A distributor fix can retain an older upstream number

Alpine lists fixed packages including 9.9_p2-r1 for 3.21, 10.0_p1-r11 for 3.22, 10.2_p1-r1 for 3.23, 10.3_p1-r2 for 3.24 and 10.6_p2-r0 for edge. Its advisory also lists fixes for older branches. [2]

Those package revisions illustrate why an upstream version number cannot be the whole acceptance criterion. In Zyel’s view, a distribution-managed server needs a distribution-aware decision: the installed package must match the vendor’s relevant fixed revision or a later supported fix. Otherwise a team can confuse a corrected backport with an unpatched installation.

A hypothetical vulnerability report might flag an older-looking banner after an update. The reviewer then needs evidence from the package and its distributor, rather than a guess based on the largest visible number. The reverse situation also deserves attention: a package recorded as updated does not answer every question about the process that is currently serving connections.

The running daemon is part of the change

Alpine explicitly requires restarting sshd after upgrading; replacing the package does not replace the running daemon. Its temporary alternative is disabling password authentication and restarting. The advisory says to verify key access and retain the existing session while testing a new login. [2]

For change planning, our recommendation is to make continued legitimate access and activation of the fix part of the same acceptance decision. Remote administration is an awkward place to discover that the only recovery plan was the session someone just closed. A maintenance record should make the intended access path and recovery responsibility clear.

The hypothetical completion evidence could identify the operating-system branch, the corrected package revision, the service restart and a successful authorized connection afterward. An exposure review can separately record which affected conditions were present before the change and what remains uncertain. These are proposed review criteria, not results from a production test performed for this article.

The broader lesson is about precision under pressure. A widely deployed component can have a narrowly defined vulnerable path, while an apparently routine update can still need a deliberate service transition. Naming both parts makes the response easier to assess: the affected login policy has been addressed, and the running daemon now reflects the chosen fix.

Sources + further reading

Source dates describe the original material. Publication above is the date this Zyel article went live.

  1. OpenSSH 10.6p2 release notesOpenSSH project · Checked Nov 10, 2026
  2. Alpine security advisory: OpenSSH password authentication bypassAlpine Linux · Checked Nov 10, 2026
  3. CISA Known Exploited Vulnerabilities catalogCISA · Checked Nov 10, 2026

Reader discussion

What would you ask next?

Questions, useful corrections and thoughtful disagreement are welcome. Comments are reviewed before they appear.

No reader comments yet.

10–2,000 characters. Plain text. Leave out passwords, private project data and personal contact details. A blank display name appears as Anonymous.

Complete the CAPTCHA before sending. Verification runs on your device.