Build regular block volumes with BlockConfigModern and submit them
through DeviceConfig::BlockCfgModern. Preserve the existing block
device attributes while also recording the host path required by the
modern block device model.
Update block volume helper call sites with the default option slice
so this change remains buildable with the modern helper signature.
Signed-off-by: Alex Lyn <alex.lyn@antgroup.com>
Switch the Cloud Hypervisor device handling path from the legacy
Block device variant to BlockModern. Keep the block device behind
its shared mutex while snapshotting the configuration for the
hotplug request, and persist the PCI path assigned by Cloud
Hypervisor back into the shared device state.
Signed-off-by: Alex Lyn <alex.lyn@antgroup.com>
Introduce build_bind_mount_options() to construct OCI mount options for
container-side bind mount, preferring volume_options over OCI options
and ensuring a bind/rbind flag is present.
And addfilter_block_storage_options() to strip bind/rbind flags from
Storage.options, preventing ENOTDIR when the agent mounts a block
device to a filesystem inside the guest.
And then Extend handle_block_volume() with a volume_options parameter
and replace hardcoded ro/empty options with the new filter+ro logic,
narrowing the function to a BlockModern-only path by removing the
DeviceType::Block branch.
Fix generate_shared_path to always create under rw/ since ro/
is a read-only bind mount of rw/ and direct creation fails with ReadOnly
FS. Switch the OCI Mount type from fs_type to KATA_MOUNT_BIND_TYPE because
the mount source is a directory (guest_path), not a block device.
Signed-off-by: Alex Lyn <alex.lyn@antgroup.com>
Move the block device type definitions shared between BlockDevice and
BlockDeviceModern — VIRTIO_BLOCK_{PCI,MMIO,CCW}/VIRTIO_PMEM constants,
KATA_*_DEV_TYPE aliases, and the BlockDeviceAio / BlockDeviceFormat
enums — from virtio_blk.rs into virtio_blk_modern.rs, so they live
alongside the BlockModern handler that will replace the legacy
BlockDevice.
Add a `format: BlockDeviceFormat` field to BlockConfigModern to match
the legacy BlockConfig. Update mod.rs re-exports accordingly.
No behavior change; pure relocation plus the new field.
Signed-off-by: Alex Lyn <alex.lyn@antgroup.com>
Extend add_block_device with num_queues and queue_size parameters
to support multi-queue virtio-blk configuration, and add BlockModern
hotplug/hotunplug handler that extracts config from
Arc<Mutex<BlockDeviceModern>>, passes the new parameters to
add_block_device, and updates pci_path on successful hotplug.
And we will use BlockDeviceModern to handle block device instead of
Legacy BlockDevice.
Signed-off-by: Alex Lyn <alex.lyn@antgroup.com>
Add BlockModern handling in DeviceManager so that device index is
correctly released both in get_device_info and in the error path.
Signed-off-by: Alex Lyn <alex.lyn@antgroup.com>
Resolve kata-monitor log level from monitor.logLevel, the chart-wide
logLevel, and debug:true, matching the precedence used by kata-deploy.
Fixes#13382
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Derive the runtime endpoint from k8sDistribution and mount the matching
host socket directory so kata-monitor can reach containerd on
non-vanilla k8s deployments.
Fixes#13378
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
We have bumped Trustee recently for the Helm chart work, but there are a
few new features that would be good to land for Kata 4.0 and CoCo
v0.22.0.
Before too long, we will start bumping Trustee to a particular version
rather than to a commit hash.
Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
Use zstd level 10 (via kata_tar_zstd) when creating static build
tarballs so CI artifacts stay under GitHub's 2GB limit without
paying the build-time cost of maximum compression.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Reuse the kubelet runtimeRequestTimeout validation for guest-pull
configurations.
Guest pull can also keep kubelet waiting during CreateContainer. Warn
when kata-deploy can see that runtimeRequestTimeout is below the same
suggested timeout used for EROFS layer conversion.
Detect both the regular guest-pull mapping and the experimental force
guest-pull configuration so nodes get the warning for either path.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Assisted-by: OpenAI <support@openai.com>
Extend the kata-deploy host-check stage with EROFS-specific
prerequisite validation.
Check that containerd supports the EROFS snapshotter, required
kernel features are available either as loaded modules or built-in
kernel configuration, and the host erofs-utils version satisfies the
minimum required by kata-deploy's configured mkfs options. Also warn
when kubelet runtimeRequestTimeout appears too low for EROFS layer
conversion, which can run during CreateContainer.
Read kubelet runtimeRequestTimeout from kubelet /configz through the
apiserver node proxy, and add the corresponding RBAC permission. Keep
fs-verity validation warning-only because user drop-ins may override
kata-deploy's default enable_fsverity setting and kata-deploy does
not yet validate the backing filesystem's fs-verity feature.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Assisted-by: OpenAI <support@openai.com>
Document operational requirements for using the EROFS snapshotter
with Kata Containers.
Call out that EROFS layer conversion can happen during
CreateContainer, not only during image pull, so kubelet
runtimeRequestTimeout may need to be increased for large images.
Also document the kernel modules needed for EROFS and dm-verity,
the erofs-utils version required by the configured mkfs options,
and the distinction between host-side fs-verity and guest-visible
dm-verity.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Assisted-by: OpenAI <support@openai.com>
Use the hotplug-returned SCSI address as Storage.source when a
container rootfs is backed by a block device.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Flesh out the quick start with the content from #13220 while keeping
mkdocs-material admonitions, definition lists, and links to installation.md.
Replace the manual Next steps nav block with in-prose cross-references, use
absolute GitHub URLs for out-of-tree source links, and update the top-level
README packaging and architecture links.
Fixes: kata-containers#12999
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Signed-off-by: LandonTClipp <lclipp@coreweave.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Switch the per-architecture default shim from the Go runtime ("qemu") to the
Rust runtime ("qemu-runtime-rs") wherever a runtime-rs build exists, and
align CI/test fallbacks with the new default.
Document the defaultShim values and Go runtime deprecation in the Helm and
config migration guides using mkdocs admonitions.
Fixes: kata-containers#9546
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Catalogue the configuration differences between the Go runtime and
runtime-rs for the QEMU hypervisor, using mkdocs admonitions and absolute
GitHub URLs for out-of-tree source links.
Address review feedback from #13236: remove the outdated msize_9p and
Summary sections, clarify that both runtimes use enable_debug while
log_level is a runtime-rs addition, and rename the architecture 4.0 page
so mkdocs does not treat it as a section index.
Co-authored-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Signed-off-by: LandonTClipp <lclipp@coreweave.com>
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Remove the RuntimeClass table in favour of prose, map `uname -m` to the
release tarball architecture names (amd64, arm64, s390x, ppc64le), mark the
Go runtime as deprecated, and point README links at installation.md.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Signed-off-by: LandonTClipp <lclipp@coreweave.com>
Assisted-by: Cursor <cursoragent@cursor.com>
src/dragonball/out is an 893 KB `cargo metadata` JSON dump that was swept
into the tree by accident in commit 1d2d2cc3d ("ci: Remove duplicated
rust-vmm dependencies"), whose only intended change was a one-line edit to
.github/dependabot.yml. Nothing references the file -- no script, CI job,
or Makefile -- so it is pure build cruft. Remove it.
Signed-off-by: Xuewei Niu <niuxuewei.nxw@antgroup.com>
At src/dragonball/, the dragonball crate's own module tree (src/) sat as a
sibling of whole separate crates (dbs_acpi/, dbs_pci/, ...). That level
mixed one crate's guts with other crates -- neither cleanly one crate nor
a clean set of crates. Move the dbs_* under src/dragonball/crates/ so the
top level is the dragonball crate plus a crates/ folder of its libraries,
matching the layout src/runtime-rs already uses.
The dbs_* were imported ~2 years ago from the standalone dragonball-sandbox
repo and kept that repo's flat layout verbatim. They are now private to
kata (unpublished, versions frozen) -- plain internal workspace members --
so no external consumers are affected and the "independently published
building blocks" rationale for the flat layout no longer holds. Pure
layout change; no code or build behavior changes.
Signed-off-by: Xuewei Niu <niuxuewei.nxw@antgroup.com>
In confidential mode, the host is not trusted and might try to create
malicious files in the guest VM. This commit changes the implementation
of the CopyFile method to use pathrs, which is designed to prevent
accidental writes to locations outside the shared directory
(/run/kata-containers/shared/containers). It also adds a detailed
description of what the method does, and test cases for intended and
malicious use cases.
This was originally considered as a mitigation candidate for
https://github.com/kata-containers/kata-containers/security/advisories/GHSA-q49m-57vm-c8cc,
but we decided to go with the simpler change to address the immediate
risk. It's still desirable to lock down CopyFile writes for defense in
depth, thus the resubmission outside the GHSA.
Signed-off-by: Markus Rudy <mr@edgeless.systems>
Use dedicated empty firmware makefile variables so the NVIDIA CPU
runtime does not inherit generic OVMF/AAVMF firmware paths.
runtime-rs already sets the correct option by default, so, no changed
needed on that front.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
The CI agent tarball is GNU-linked (USE_DEVMAPPER=yes) and does not run
reliably on Alpine/musl initrd.
Use Ubuntu noble for x86_64 and aarch64 vanilla initrd, matching the
other initrd variants, with kata-agent as /sbin/init.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Extend the post-delete RBAC cleanup Job to remove job-mode dispatcher
identities (ServiceAccount, Role, RoleBinding, ClusterRole, and
ClusterRoleBinding) so reinstalls are not blocked by leftovers.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Update confidential_kbs.sh to use the renamed Helm chart knobs
for IBM SEL credentials: `as.ibmse.*` → `as.verifier.se.*`.
Bump coco-trustee version and all associated image tags from
80a3c55e to b6c8dc2b accordingly.
Signed-off-by: Hyounggyu Choi <Hyounggyu.Choi@ibm.com>
CI doesn't yet enable bundled Postgres for the Trustee Helm chart, but
Helm still tried to pull the Bitnami subchart from Docker Hub OCI during
dependency build, hitting unauthenticated rate limits on shared runners.
Once we switch to using Postgres in the tests, we will need to drop this
workaround and revisit this issue.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
The job dispatcher now GETs each in-flight Job by name when tracking
progress, so it no longer needs the list verb on batch/jobs. Trim the
dispatcher Role to match.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Assisted-by: Cursor <cursoragent@cursor.com>
Add a try-kata-nvidia-cpu.values.yaml example for the CPU-only
NVIDIA runtime classes and make the Kubernetes CI Helm helper use it
for qemu-nvidia-cpu handlers.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Assisted-by: OpenAI Codex <codex@openai.com>
The qemu-nvidia-cpu runtime classes do not provide GPU
passthrough, so do not list them as supported by the NVIDIA
GPU test runner.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Assisted-by: OpenAI Codex <codex@openai.com>
dmverity_mode=auto can preserve existing EROFS layers without
dm-verity metadata, which lets mixed-mode nodes keep running with
partially unprotected image layers.
Use dmverity_mode=on when kata-deploy enables EROFS dm-verity so
missing metadata is reported instead of silently accepted.
Also tighten the Kubernetes probe. Instead of checking for at least one
dm-verity line and one EROFS dm mount in dmesg, derive the expected
workload layer count from the root overlay lowerdirs and require matching
kata-verity devices and dm-backed EROFS mounts.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
Assisted-by: OpenAI Codex <codex@openai.com>
