* fix(driver-mxc): pass gateway name to mocked backend
Refs #4191
Signed-off-by: Philippe Martin <phmartin@redhat.com>
* fix(windows): use native executable paths in network test fixtures
Signed-off-by: Philippe Martin <phmartin@redhat.com>
---------
Signed-off-by: Philippe Martin <phmartin@redhat.com>
Remove blanket review routing while preserving the explicit owners of the vouch list, workflows, actions, approval helpers, and SBOM check already on main.
Signed-off-by: Jim Meyer <jimeyer@nvidia.com>
* test(tmachine): add K3s PostgreSQL and mTLS conformance installer
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* fix(tmachine): reach the HA TLS gateway through its ClusterIP
Drop the port-forward restart removed by the ClusterIP change, resolve the Service DNS name covered by the server certificate to the ClusterIP, bound each authenticated API wait attempt, and collect diagnostics from every gateway Pod.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* ci(tmachine): run K3s PostgreSQL and mTLS conformance
Add the k3s-ha-tls installer to every core conformance matrix so Branch E2E, Release Dev, Release Tag, and manual Integration Tests cover three gateway replicas sharing PostgreSQL behind mTLS.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* refactor(tmachine): rename k3s-ha-tls installer to k3s-ha
TLS is part of the supported HA configuration once gateway replicas share PostgreSQL, so the installer name only needs to distinguish HA from the single-replica k3s lane.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* refactor(tmachine): install K3s HA in a single Helm rollout
Add an optional Helm values overlay to the K3s installer so the k3s-ha
installer stages PostgreSQL first and installs the three-replica TLS
gateway directly, instead of installing a plaintext gateway and upgrading.
Drop the redundant unauthenticated-users override and the install-time
API wait, which the conformance playbook already performs. Share K3s
installer inputs in the tmachine configuration.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* fix(tmachine): make K3s HA staging rerunnable and wait for PostgreSQL
Apply the OpenShell namespace instead of creating it so the staging play
tolerates an existing namespace, and wait for the PostgreSQL rollout so
gateway Pods do not start crash-looping against a missing database.
Reflow and pluralize the K3s lane notes in CI.md.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
---------
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
The runtimes page examples omit the [openshell] version = 2 table, so a
snippet copied into a new gateway.toml fails preflight with
missing_version. Add the table to the first example and state that the
remaining examples show only the tables they change.
Closes#3881
Signed-off-by: fede-kamel <fkamelhar@gmail.com>
Spell out that auto-closed PRs keep their discussion and can be reopened or replaced, separate DCO sign-off from cryptographic commit signing, and say to mark PR checklist items complete only after verification.
Signed-off-by: politerealism <burdcat17@gmail.com>
* feat(python): expose sandbox status conditions in the SDK
Add SandboxConditionRef and surface status.conditions on SandboxStatusRef and SandboxRef. Callers need condition reasons and transition times to time terminal transitions and explain failures that happen before the main process starts, without reading the raw proto.
Signed-off-by: Max Dubrinsky <mdubrinsky@nvidia.com>
* test(python): use real gateway condition values in conditions fixtures
Signed-off-by: Max Dubrinsky <mdubrinsky@nvidia.com>
---------
Signed-off-by: Max Dubrinsky <mdubrinsky@nvidia.com>
The stop path wrote Stopping and Stopped with a single compare-and-swap.
In HA deployments every replica applies compute driver watch events for
the terminating runtime without the per-replica lifecycle gate, so a
watch write landing between the re-read and the CAS failed the stop RPC
with "update sandbox lifecycle failed due to concurrent modification"
and left the sandbox in Stopping.
Retry both writes on PersistenceError::Conflict up to
STOP_PHASE_CAS_RETRY_LIMIT, re-reading the sandbox on each attempt. The
Stopping write re-runs the stop preconditions and the name check; the
Stopped write proceeds only while the record is still Stopping, returns
an already Stopped record as-is, and aborts on any other phase. Gateway
startup recovery of retained Stopping sandboxes now shares the same
helper instead of writing against its pre-driver snapshot.
Signed-off-by: Piotr Mlocek <pmlocek@nvidia.com>
* feat(providers): support stable credential placeholders
Add credentials[].stable_placeholder so a running client keeps using an externally rotated static credential through the reference it already holds. Squashed replay of the earlier four commits onto upstream main 71c3cd957; re-derives the frozen schema fingerprints and adapts the provider environment revision test to the added capability argument.
Signed-off-by: Shiju <shiju@nvidia.com>
* fix(e2e): apply stable-placeholder supervisor image override
Use ManagedGateway's per-process supervisor image override so the fixture
CA reaches the supervisor when runtime image environment variables take
precedence over gateway TOML. Restore the wrapper's original image after
each test, following the provider-readiness fixture.
Signed-off-by: Shiju <shiju@nvidia.com>
* test(server): update public RPC schema fingerprint after merging main
Signed-off-by: Shiju <shiju@nvidia.com>
---------
Signed-off-by: Shiju <shiju@nvidia.com>
* test(k3s): connect tmachine clients directly to ClusterIP
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* test(conformance): wait for a connected gateway on every installer
Replace the K3s-specific boot readiness hook with one shared gateway wait in the conformance playbook. Skip conformance and collect the existing diagnostics when the gateway never connects. Discover the gateway Service port with its ClusterIP, and capture both iptables backends in K3s diagnostics.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
* test(tmachine): centralize gateway diagnostics and trim K3s changes
Move failure diagnostics into one openshell_diagnostics role that collects the systemd gateway journal and, when K3s is installed, the K3s diagnostics. Drop install-time diagnostics so the K3s installer change stays minimal, revert unrelated skill edits, and shorten the CI.md note.
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
---------
Signed-off-by: Matthew Grossman <mgrossman@nvidia.com>
Adds the OpenShell / Maintainer Approval status check. It passes only when
someone listed in MAINTAINERS.md has approved the pull request, and it
notifies no one: a required status check is the only native primitive that
enforces who must approve without requesting a review from them.
The job runs on pull_request_review and merge_group with read-only
permissions and posts nothing anywhere; its exit code is the check result.
Both MAINTAINERS.md and the decision helper are read from main via a
sparse checkout, never from the pull request ref, so a contributor cannot
add themselves to the list or rewrite the decision logic and self-approve.
Because pull_request_review runs the pull request's own copy of the
workflow file, CODEOWNERS now covers the whole enforcement surface --
workflows, actions, CODEOWNERS itself, the zizmor config, MAINTAINERS.md,
and the enforcement scripts -- naming the outside collaborators explicitly
alongside the team, since org policy keeps them off GitHub teams. The
ruleset's native code owner requirement is what blocks a merge that
disables the gate, regardless of what the tampered check reports.
Also adds a change-alert workflow that comments the approver-set delta on
pull requests touching MAINTAINERS.md, and documents both in CONTRIBUTING.
Signed-off-by: Jim Meyer <jimeyer@nvidia.com>
* feat(snap): run gateway as a user service by default
Replace the existing system `gateway` service with separate
`user-gateway` and `system-gateway` services. For new installs, only the
`user-gateway` service is enabled. For existing installs, only the
`system-gateway` service is enabled.
Existing installs continue to have a one-time migration which removes
legacy insecure configurations. It is up to `install.sh` or users to
manually copy mTLS credentials from the root-owned `$SNAP_COMMON/tls` to
the invoking user's OpenShell snap directory. This is explained in the
snap description in `snapcraft.yaml`, visible in the Snap Store listing
and via `snap info openshell`.
Service enablement is now managed by a new snap configuration option
named `gateway-mode`, so users can switch from one mode to another via
e.g. `sudo snap set openshell gateway-mode=user`. The `install` and
`post-refresh` hooks select which service to start by setting this mode.
If the `gateway-mode` is already set, then we know the one-time
migration has already taken place.
Signed-off-by: Oliver Calder <oliver.calder@canonical.com>
* fix(snap): simplify the snap gateway wrapper
Signed-off-by: Oliver Calder <oliver.calder@canonical.com>
* fix(snap): use `snap start --user` when running as non-root
Signed-off-by: Oliver Calder <oliver.calder@canonical.com>
* fix(snap): ensure install.sh waits for the correct docker daemon
Signed-off-by: Oliver Calder <oliver.calder@canonical.com>
* feat(snap): document how to safely switch between gateway modes
New installs run the gateway as a user service by default, while
upgrades of existing installs from prior to the user gateway continue to
use the existing system gateway. The user and system gateway are
fundamentally different services and there is no way to migrate state
between them.
However, it may be desirable for a user to run the gateway as a system
service after a fresh install. To support this, this commit adds
documentation for how to safely transition from the user gateway mode to
the system gateway mode, and vice versa.
Also, clarifies when it is necessary to manually copy TLS credentials
from the system gateway state directory to the user's home directory:
when upgrading a pre-TLS snap installation, or when switching from user
to system gateway mode. When upgrading an existing install with an HTTPS
gateway running as a system service, copying credentials again is not
necessary.
Signed-off-by: Oliver Calder <oliver.calder@canonical.com>
---------
Signed-off-by: Oliver Calder <oliver.calder@canonical.com>
## Summary
Bump the pinned libkrunfw from v5.6.1 to v5.6.2, the latest upstream
release. The guest kernel stays on Linux 6.12.109 and the ABI stays at 5.
## Changes
- libkrunfw: v5.6.1 (617938cf) -> v5.6.2 (f14ac64c)
Upstream changes in v5.6.2: swap, zram (zstd), MGLRU and PSI enabled in
the aarch64 and x86_64 kernel configs, PAT/MTRR enabled on x86_64 for
write-combining mappings, and native macOS kernel builds. The Linux build
path of the Makefile ($(KERNEL_SOURCES) and kernel.c targets) used by
build-libkrun.sh is unchanged.
Signed-off-by: Florent Benoit <fbenoit@redhat.com>
## Summary
Bump the pinned libkrun from v1.19.4 to v1.19.6, the latest upstream
release. v1.19.5 includes libkrun/libkrun@d2d8dd6, which removes refused
vsock connections without waiting for the 5 s reaper, so VM sandbox
startup no longer stalls on "Waiting for VM supervisor".
## Related Issue
Closes#4259
## Changes
- libkrun: v1.19.4 (728df812) -> v1.19.6 (227b2de6)
- Update LIBKRUN_REF in pins.env and the fallback defaults in
build-libkrun.sh and build-libkrun-macos.sh
- libkrunfw stays on v5.6.1
Other notable upstream fixes (v1.19.5): vsock proxy release after clean
shutdown, macOS virtio-fs xattr handling, x86_64 initrd kept below the
MMIO gap, PIT missed-tick reinjection disabled, macOS VMs with >64GB
RAM. v1.19.6 adds an API to set the render server fd. No API removals.
## Testing
- libkrun v1.19.6 builds on macOS arm64 with `make BLK=1 NET=1`
Signed-off-by: Florent Benoit <fbenoit@redhat.com>
Reject host paths and malformed CDI selectors for Docker and Podman
using shared qualified-name validation. Preserve the generic GPU
request checks for drivers using other identifier formats, including
VM PCI addresses.
Validate explicit and resolved CDI selectors, cover vendor-agnostic
names and rejected paths, and update the related documentation.
Signed-off-by: Adrien Langou <alangou@nvidia.com>
Signed-off-by: Evan Lezar <elezar@nvidia.com>
Evan Lezar