Using `-E` preserves environment variables, except for PATH, so PATH has to be
manually set to match the current environment.
I removed env-vars that were redundant (such as `GOPATH=$GOPATH`), which should
be handled by `-E`.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
cri-tools is hardcoded to use images which are broken
within their registry. Disable the tests to unblock
CI until fixed.
Signed-off-by: Derek McGowan <derek@mcg.dev>
This allows us to dig more details out of test runs and maintain a
better history.
For this we can use `gotestsum`, which is a utility that wraps `go test`
so that it outputs test2json (go's format) and output junit (a format
more easily imported into other systems).
The PR makes it possible to override the Makefile's use of `go test` to
use any other command tto executet the test. For CI we'll use `gotestsum
--`, where `gotestsum` expects everything after the `--` to be flags for
`go test`.
We then use environment variables to configure `gotestsum` (e.g.
`GOTESTSUM_JUNITFILE` is an env var accepted by `gotestsum`).
For cri tests, the test suite supports outputing test results to a
directory, these are in junit format already. The file is not named
properly just because the code that creates it (in ginkgo) is not
configured well. We can fix that upstream to give us a better name...
until then I'm keeping those results in a separate dir.
A second workflow is also added so the test results can be summed up and
a report added to the workflow run. The 2nd workflow is required for
this since PR runs do not have access to do some of this due to safety
reasons
(https://securitylab.github.com/research/github-actions-preventing-pwn-requests/)
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Though we don't officially support Apple platform, we should
at least run unit tests to make sure things are not broken.
Signed-off-by: Maksym Pavlenko <pavlenko.maksym@gmail.com>
The integration test times have slightly increased and are often
hitting the 25 minutes timeout. This increases to give more room
but still keeps it low enough to catch regressions in tests
causing longer than expected execution.
Signed-off-by: Derek McGowan <derek@mcg.dev>
While the issue hasn't been fixed in the kernel yet, we can workaround
the issue by not using overlayfs snapshotter.
The newly added step runs all tests that match /TestCheckpoint/.
So, TestCRWithImagePath has been renamed to match the regexp.
Fixes#3930.
Signed-off-by: Kazuyoshi Kato <katokazu@amazon.com>
To investigate issues like #4969, it would be helpful to understand
the status of the VM at the end.
Signed-off-by: Kazuyoshi Kato <katokazu@amazon.com>
Without a name the logs use a carriage return followed by the long
comment as the name of the job step which is messy when working with the
actions API/logs.
Signed-off-by: Phil Estes <estesp@amazon.com>
Since both cri-containerd and runC runtime are widely used, the relevent
information should include runC and CRI configuration when file bug.
Signed-off-by: Wei Fu <fuweid89@gmail.com>
The CI host was probably updated recently and the permission bits of the
directory was changed.
Fix 5077
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
We have enough failures these days; getting timed out when tests are
almost done is the last thing we need :)
On avg. the Linux integration tests are taking 15-17 min, but sometimes
they end up at 20 or a bit over and get canceled. I've seen rare cases
where the Vagrant setup+build+test runs gets very close to 40 min as
well.
Signed-off-by: Phil Estes <estesp@amazon.com>
https://github.com/actions/setup-go/tree/v2.1.3#v2
The V2 offers:
- Adds GOBIN to the PATH
- Proxy Support
- stable input
- Bug Fixes (including issues around version matching and semver)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
download.fedoraproject.org gives HTTP 404 at times,
breaking the CI. Let's give it another chance.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Add `.github/ISSUE_TEMPLATE/config.yml` to clarify where is the right place (in the most cases) to ask questions.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
CI was timing out after 15 minutes on the crun tests; extending
the timeout to 20 minutes (we can make it shorter again if we know
the exact time it takes to run)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Noticed this in the CI output:
Requested golangci-lint 'v1.29', using 'v1.29.0', calculation took 7969ms
Installing golangci-lint v1.29.0...
Downloading https://github.com/golangci/golangci-lint/releases/download/v1.29.0/golangci-lint-1.29.0-darwin-amd64.tar.gz ...
Using nearly 8 seconds to convert v1.29 to v1.29.0 seems a bit long,
so hard-coding to the full version to speedup CI somewhat.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
With container-selinux policy updated to 2.145+ (the default for Fedora 32+) we
can enable SELinux=Enforcing mode in the CI workflow and pass all integration
and CRI tests except one, see https://github.com/containerd/containerd/issues/4460,
which has been marked as skipped.
Tested locally with:
- SELINUX=Enforcing vagrant up --provision-with=shell,selinux,test-integration
- SELINUX=Enforcing vagrant up --provision-with=shell,selinux,test-cri
Signed-off-by: Jacob Blain Christen <jacob@rancher.com>
These checks had to be disabled to get the CRI merge completed.
Now these should be added back.
After CRI merge, more time for lint is needed on mac.
Signed-off-by: Derek McGowan <derek@mcg.dev>
Add installation scripts needed to generate CRI + CNI tar package on
release.
Update Github action release script to generate CRI release tarballs for
Linux and Windows.
Signed-off-by: Derek McGowan <derek@mcg.dev>
`vagrant up` will build and install containerd and all dependencies,
setting up proper SELinux contexts on the runc and containerd binaries.
The VM is configured to be SELinux Enforcing by default but this gets
changed during various CI passes via a matrix param to Disabled and
Permissive before running tests. I have an open PR to fix the
container-selinux policy for containerd at
https://github.com/containers/container-selinux/pull/98 which once
accepted we will want to update the CI matrix to use Enforcing mode
instead of Permissive.
All tests currently pass in SELinux permissive mode with containerd
configured with `enable_selinux=true`. To see which tests are failing
with SELinux enforcing and an already spun up VM:
`SELINUX=Enforcing vagrant up --provision-with=selinux,test-cri`
To test SELinux enforcing in a new VM:
`vagrant destroy -force; SELINUX=Enforcing vagrant up --provision-with=shell,selinux,test-cri`
The `selinux` shell provisioner, parameterized by the SELINUX envvar,
will configure the system as you would expect, with the side effect that
containerd is configured with `enable_selinux=true` via
`/etc/containerd/config.toml` for Permissive or Enforcing modes and
`enable_selinux=false` when SELINUX=Disabled.
Provided that virtualization is suported, this Vagrantfile and provisioners
make it easy to test containerd/cri for conformance under SELinux on
non-SELinux systems.
Signed-off-by: Jacob Blain Christen <jacob@rancher.com>
Recent changes removed the need for libseccomp-dev when building
containerd. The btrfs tools package is already installed on GH Actions
runners and was already a no-op so the whole step can be removed.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
This will ensure that we can always get the current Windows OS build
version, without being put into Windows 8 compatibility mode.
Signed-off-by: Paul "TBBle" Hampson <Paul.Hampson@Pobox.com>
We are going to deprecate shim v1 (#4365), but it is still early to
disable the tests for them
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
GH runners now have a systemd-run containerd running on the standard
socket, impacting the CRI test's expectation of our CI-built containerd
running there.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
The previous names were confusing because v1 may look like shim API v1 (io.containerd.runtime.v1.linux").
Also, v2 may look like cgroup v2.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
- always apt-get update before installing packages
- move to tagged official create_release action
The official GH create_release action now has support for body text from
file.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
[BeforeEach] [k8s.io] Security Context
/home/runner/work/containerd/containerd/src/github.com/kubernetes-sigs/cri-tools/pkg/framework/framework.go:50
W0624 12:26:28.532644 30569 util_unix.go:103] Using "/var/run/containerd/containerd.sock" as endpoint is deprecated, please consider using full url format "unix:///var/run/containerd/containerd.sock".
W0624 12:26:28.532700 30569 util_unix.go:103] Using "/var/run/containerd/containerd.sock" as endpoint is deprecated, please consider using full url format "unix:///var/run/containerd/containerd.sock".
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This will check that the tag is signed and then checkout the tag, build
official binaries, sha256sum the tarball, and upload those assets to the
release, officially generating a release in GitHub from the signed tag.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
- Remove dependency on libcontainer/system
- Get rid of socat for port forwarding
- Roll docker/distribution back to latest (v2.7.1) release
Now that 901bcb2231 was merged in containerd,
we no longer depend on the ParseDockerRef utility from docker/distribution,
so we can safely roll back to the latest release for this dependency.
- vendor: kubernetes v1.18.2
Fix client watch reestablishment handling of client-side timeouts
- Add config flag to default empty seccomp profile
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
omit sudo when EUID eqto 0
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
omit sudo when EUID eqto 0
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
omit sudo when EUID eqto 0
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
use gosu to omit sudo in GA
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
use gosu to omit sudo in GA
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
set working-dir for <<Setup gosu>> step
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
fix job permissions
Signed-off-by: fahedouch <fahed.dorgaa@gmail.com>
Integration tests were running with latest Go release rather than the
version used everywhere else. Also, we don't need to install protoc from
tarball and also apt-get the package for Ubuntu when used as a
dependency for criu build.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
Also few changes to names of workflow jobs. Integration tests can get
canceled without full completing at 10 minutes. Make sure we only use
checkout v2 action.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
full diff: https://github.com/golang/go/compare/go1.13.6...go1.13.7
go1.13.7 (released 2020/01/28) includes two security fixes. One mitigates
the CVE-2020-0601 certificate verification bypass on Windows. The other affects
only 32-bit architectures.
https://github.com/golang/go/issues?q=milestone%3AGo1.13.7+label%3ACherryPickApproved
- X.509 certificate validation bypass on Windows 10
A Windows vulnerability allows attackers to spoof valid certificate chains when
the system root store is in use. These releases include a mitigation for Go
applications, but it’s strongly recommended that affected users install the
Windows security update to protect their system.
This issue is CVE-2020-0601 and Go issue golang.org/issue/36834.
- Panic in crypto/x509 certificate parsing and golang.org/x/crypto/cryptobyte
On 32-bit architectures, a malformed input to crypto/x509 or the ASN.1 parsing
functions of golang.org/x/crypto/cryptobyte can lead to a panic.
The malformed certificate can be delivered via a crypto/tls connection to a
client, or to a server that accepts client certificates. net/http clients can
be made to crash by an HTTPS server, while net/http servers that accept client
certificates will recover the panic and are unaffected.
Thanks to Project Wycheproof for providing the test cases that led to the
discovery of this issue. The issue is CVE-2020-7919 and Go issue golang.org/issue/36837.
This is also fixed in version v0.0.0-20200124225646-8b5121be2f68 of golang.org/x/crypto/cryptobyte.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
