On this page (6 sections)

CVE-2026-130780 is a critical path traversal flaw in Incus, the container and virtual machine manager. According to the NVD record, versions before 7.6.0 trust names supplied by a migration peer, which lets a malicious peer write outside the directories Incus intends to use.

If you run Incus hosts that accept instance or volume migrations from other servers, or that use dir storage pools, this is worth checking today. The research available so far is limited to the NVD entry and a linked GitHub advisory, so some details below are marked as unconfirmed.

What happened

The NVD description says Incus before 7.6.0 does not validate the names of dependent volumes and snapshots that arrive in the migration offer header. A migration is when one Incus server sends an instance or storage volume to another. If the sending side is malicious or compromised, it can put crafted names in that header.

The record states that an attacker controlling a migration source can create root-owned volumes containing attacker-chosen content at arbitrary paths on the host. The upstream project published its own GitHub security advisory, which is the primary reference for this issue.

Technical details in plain terms

Path traversal (CWE-22) happens when software builds a file path from untrusted input without checking that the result stays inside an intended folder. Values such as parent-directory references can then point outside it. Here the untrusted input is the names of dependent volumes and snapshots in the migration header, and the files are created as root-owned.

The NVD text lists three outcomes:

  • Host compromise on systems that use dir storage pools, because attacker content can land at arbitrary host paths.
  • A bypass of restricted.storage-pools.access, the setting meant to limit which storage pools a restricted project may use.
  • A crash of the Incus daemon.

The sources do not describe a specific exploitation chain, a required authentication level beyond controlling a migration source, or proof-of-concept code, and this article does not speculate about them.

Who is realistically exposed

Exposure depends on whether your host will ever receive a migration from a source you do not fully trust. Higher-risk setups include:

  • Hosts that accept migrations from other servers, clusters or users, including shared or multi-tenant environments.
  • Hosts using dir storage pools, which the record names as directly compromisable.
  • Environments that rely on restricted.storage-pools.access to separate projects or tenants.
  • Hosts where a migration source could be an account or machine that has been compromised.

A single-user host that never receives migrations has a smaller practical exposure, but the vulnerable code is present in every version before 7.6.0, so upgrading is still the right fix.

How to check your exposure

Start with the installed version. This read-only command, run on each Incus host, prints the client and server versions. Anything below 7.6.0 is in the affected range.

Show Incus client and server versions (run on each host)Shell
incus version

Next, list your storage pools and note which use the dir driver. This read-only command shows each pool and its driver.

List storage pools and their driversShell
incus storage list

Finally, review which clients and servers are trusted to talk to your hosts, since a migration source normally needs to be trusted or hold a valid token.

List trusted certificates/clients for this serverShell
incus config trust list

What to do now

Caution. Upgrading Incus can restart the daemon. Check how your instances behave across a daemon restart and schedule the change in a maintenance window, especially on production hosts.
  1. Run incus version on every host and record which are below 7.6.0.
  2. Upgrade affected hosts to Incus 7.6.0 or later using the package source you installed from, and follow the project's upgrade notes.
  3. Until you can upgrade, stop accepting migrations from sources you do not fully control, and do not start migrations from untrusted peers.
  4. Review trusted clients with incus config trust list and remove any you do not recognize.
  5. If you use dir storage pools or restricted projects, treat hosts that received migrations from untrusted sources as suspect and inspect them for unexpected files owned by root.
  6. Read the upstream advisory for the authoritative fixed-version and workaround details.

After upgrading, confirm the new version with the same command.

Confirm the upgrade (run on each host)Shell
incus version

Detection and monitoring

The research gives no indicators of compromise, so there is no signature to search for. Practical steps are to review Incus daemon logs around past migrations for errors or crashes, since the record notes a daemon crash is one possible outcome, and to look for recently created root-owned files or directories in places that should not change, such as system configuration directories.

Treat an unexplained Incus daemon crash coinciding with a migration as a reason to investigate rather than just restart. If you find signs of tampering, follow a standard recovery process such as the one in Website hacked: what to do.

If the services running in your instances are public websites, keep an eye on their certificates and DNS for unexpected changes after any host incident. Webforta's monitoring covers certificate, DNS and uptime changes, though it does not look inside your hypervisor hosts.

Only one independent publisher covers this specific CVE so far, so check the upstream advisory for updates.

Sources

This article is based on the public sources below. Details can change as vendors and researchers publish more; check the original advisories before acting on version numbers.

  1. CVE-2026-130780: Incus before 7.6.0 contains a path traversal vulnerability that allows malicious migration peers to write outside intended directories by supplying…nvd.nist.gov ·
  • Incus
  • CVE-2026-130780
  • path traversal
  • container security
  • virtualization