That's not needed anymore, as we've switched to using ORAS and an OCI
registry to cache the artefacts.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
This is something that was done by our Jenkins jobs, but that I ended up
missing when writing d0c257b3a7.
Now, let's also add the sha256sum to the cached artefact, and in a
coming up PR (after this one is merged) we will also start checking for
that.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
In the previous series related to the artefacts we build, we've
switching from storing the artefacts on Jenkins, to storing those in the
ghcr.io/kata-containers/cached-artefacts/${artefact_name}.
Now, let's take advantage of that and actually use the artefacts coming
from that "package" (as GitHub calls it).
NOTE: One thing that I've noticed that we're missing, is storing and
checking the sha256sum of the artefact. The storing part will be done
in a different commit, and the checking the sha256sum will be done in a
different PR, as we need to ensure those were pushed to the registry
before actually taking the bullet to check for them.
Fixes: #7834 -- part 2
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The list of tests which require a bigger VM instance is:
* k8s-number-cpus.bats -- failing on all CIs
* k8s-parallel.bats -- only failing on the cbl-mariner CI
* k8s-scale-nginx.bats -- only failing on the cbl-mariner CI
We'll keep those disabled while we re-work the logic to **only run
those** in a bigger (and more expensive) VM instance.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Let's push the artefacts to ghcr.io and stop relying on jenkins for
that.
Fixes: #7834 -- part 1
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Right now this is not used, but it'll be used when we start caching the
artefacts using ORAS.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
ORAS is the tool which will help us to deal with our artefacts being
pushed to and pulled from a container registry.
As both the push to and the pull from will be done inside the
kata-deploy binaries builder container, we need it installed there.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Without setting the cpu limit / request to 1, we can make this test run
in a smaller VM instance without any issue.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
As the environment variables are now being passed down from the GitHub
Actions, let's make sure they're exposed to the container used to build
the kata-deploy binaries, and during the build process we'll be able to
use those to log in and push the artefacts to the OCI registry, using
ORAS.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
We do the build of our artefacts inside a container image, and we need
to expose some env vars to the container so ORAS can be used there to
push the artefacts we want to cache to ghcr.io.
The env vars we're exposing are:
* ARTEFACT_REGISTRY: The registry where we're going to save the
artefacts.
* ARTEFACT_REGISTRY_USERNAME: The username to log in to the registry, as
ORAS does not use the same json file used by docker.
* ARTEFACT_REGISTRY_PASSWORD: The pasword to log in to the the registry,
as the ORAS does not use the same json file used by docker.
* TARGET_BRANCH: The target branch, which will be part of the tag of the
artefact, as we may end up caching the artefacts for both main and
stable branches.
Fixes: #7834 -- part 0
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
We need a very recent L2 guest kernel to fix all the bugs that occur in nested
virtualization.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
cloud hypervisor does not emulate pcie switches or pci bridges, so we need to
accept a lonely device.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
It is fine to start a VM with the disk image without syncing it as we now run
the test in an ephemeral Azure instance.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
Sometimes the test gets stuck running commands in the container - need to
investigate why later.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
Cloud Hypervisor exposes a VIRTIO_IOMMU device to the VM when IOMMU support is
enabled. We need to add it to the whitelist because dragonball uses kernel
v5.10 which restricted VIRTIO_IOMMU to ARM64 only.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
by enabling IOMMU on the default PCI segment. For hotplug to work we need a
virtualized iommu and clh exposes one if there is some device or PCI segment
that requests it. I would have preferred to add a separate PCI segment for
hotplugging vfio devices but unfortunately kata assumes there is only one
segment all over the place. See create_pci_root_bus_path(),
split_vfio_pci_option() and grep for '0000'.
Enabling the IOMMU on the default PCI segment requires passing enabling IOMMU on
every device that is attached to it, which is why it is sprinkled all over the
place.
CLH does not support IOMMU for VirtioFs, so I've added a non IOMMU segment for
that device.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
There is no way for this branch to be hit, as port is only set when it is
different than config.NoPort.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
These test cases shows which options are valid for CLH/Qemu, and test that we
correctly catch unsupported combinations.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
The only supported options are hot_plug_vfio=root-port or no-port.
cold_plug_vfio not supported yet.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
hot_plug_vfio needs to be set to root-port, otherwise attaching vfio devices to
CLH VMs fails. Either cold_plug_vfio or hot_plug_vfio is required, and we have
not implemented support for cold_plug_vfio in CLH yet.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
tdp_mmu had some issues up until around Linux v6.3 that make it work
particularly bad when running nested on Hyper-V. Reload the module at the start
of the test and disable the tdp_mmu param.
Gather debug info at the end of the test to make it easier to figure out what
went wrong. This uses github actions group syntax so that each section can be
collapsed.
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
- reduce memory and cpu usage to fit in a D4s_v5
- source correct lib
- mount workspace from 9p
- disable cpu mitigations for speed
- drop unused commands and variables
- install containerd
- install kata from built artifacts
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
This imports the vfio test scripts github.com/kata-containers/tests. The test
case doesn't work yet but doing the changes in a separate commit will make it
easier to track the changes. The only change in this commit is renaming
vfio_jenkins_job_build.sh -> vfio_fedora_vm_wrapper.sh
Fixes: #6555
Signed-off-by: Jeremi Piotrowski <jpiotrowski@microsoft.com>
Otherwise we'll fail to configure kata-containers in the `install-kata`
step.
This is mostly needed because the nerdctl-full tarball doesn't provide a
contaienrd configuration, just the binary, as contaienrd does not
actually require a configuration file to run with the default config.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
TIL that the Azure VMs we use are created without an explicit outbund
connectivity defined.
This leads us to issues using `ping ...` as part of our tests, and when
consulting Jeremi Piotrowski about the issue he pointed me out to two
interesting links:
* https://learn.microsoft.com/en-us/azure/virtual-network/ip-services/default-outbound-access
* https://learn.microsoft.com/en-us/archive/blogs/mast/use-port-pings-instead-of-icmp-to-test-azure-vm-connectivity
For your own sanity, do not read the comments, after all this is
internet. :-)
Anyways, the suggestion is to use nping instead, which is provided by
the nmap package, so we can explicitly switch to using the tcp port 80
for the ping. With this in mind, I'm switching the image we use for the
test and using one that provided nping as a possible entry point, and
from now on (this part of) the tests should work.
Fixes: #7910
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
TIL that the Azure VMs we use are created without an explicit outbund
connectivity defined.
This leads us to issues using `ping ...` as part of our tests, and when
consulting Jeremi Piotrowski about the issue he pointed me out to two
interesting links:
* https://learn.microsoft.com/en-us/azure/virtual-network/ip-services/default-outbound-access
* https://learn.microsoft.com/en-us/archive/blogs/mast/use-port-pings-instead-of-icmp-to-test-azure-vm-connectivity
For your own sanity, do not read the comments, after all this is
internet. :-)
Anyways, the suggestion is to use nping instead, which is provided by
the nmap package, so we can explicitly switch to using the tcp port 80
for the ping. With this in mind, I'm switching the image we use for the
test and using one that provided nping as a possible entry point, and
from now on (this part of) the tests should work.
Fixes: #7910
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
