This is mainly/initially to support use of Cirrus-CI
in https://github.com/containers/buildah since that setup
re-uses the VM images from this project. However, it also
opens doors here, if libpod ever needs/wants to do things
with a dedicated storage device and/or storage-drivers.
Signed-off-by: Chris Evich <cevich@redhat.com>
The release-task ***must*** always execute last, in order to guarantee a
consistent cache of release archives from dependent tasks. It
accomplishes this by verifying it's task-number matches one-less than
the total number of tasks. Previous to this commit, a YAML anchor/alias
was used to avoid duplication of the dependency list between 'success'
and 'release'
However, it's been observed that this opens the possibility for
'release' and 'success' tasks to race when running on a PR. Because
YAML anchor/aliases cannot be used to modify lists, duplication is
required to make 'release' actually depend upon 'success'.
This duplication will introduce an additional maintenance burden.
Though when adding a new task, it's already very easy to forget to
update the 'depends_on' list. Assist both cases by the addition
unit-tests to verify ``.cirrus.yml`` dependency contents and structure.
Signed-off-by: Chris Evich <cevich@redhat.com>
It's been observed on several occasions, some tests fail in git clones
with a "cannot find ref" type error. Especially in the depth=1 cases.
Since there's really only one place where limiting the depth makes sense
(build-each-commit), simply remove all the other limits.
Signed-off-by: Chris Evich <cevich@redhat.com>
clean up some final linter issues and add a make target for
golangci-lint. in addition, begin running the tests are part of the
gating tasks in cirrus ci.
we cannot fully shift over to the new linter until we fix the image on
the openshift side. for short term, we will use both
Signed-off-by: baude <bbaude@redhat.com>
The 'image_prune' task only runs on master, post-merge and
is currently failing for all builds. This is because it
references a non-existent image. The person with access
to add/enable this image is on PTO. Fix this by temporarily
using a hand-built image until an automatic build can be added.
Signed-off-by: Chris Evich <cevich@redhat.com>
Cirrus-CI automatically sets `$CIRRUS_BASE_BRANCH` during PR testing.
This is used for the `build_each_commit` task, in order to compute the
commit-chain properly. However, prior to this commit and after a PR
merges, the post-merge `build_each_commit` task would fail with
something similar to:
```
make build-all-new-commits GIT_BASE_BRANCH=origin/$CIRRUS_BASE_BRANCH |& ${TIMESTAMP}
[12:28:59] START - All [+xxxx] lines that follow are relative to right now.
[+0000s] # Validate that all the commits build on top of origin/
[+0000s] git rebase origin/ -x make
[+0000s] fatal: invalid upstream 'origin/'
[+0000s] make: *** [Makefile:426: build-all-new-commits] Error 128
[12:28:59] END - [+0000s] total duration since START
Exit status: 2
```
This is because `$CIRRUS_BASE_BRANCH` is undefined when CI runs against
a branch (by design). This commit fixes the problem by referring to
`$DEST_BRANCH` instead. This variable must always point at the intended
destination branch for testing, and so can be used in this context as
well.
Also updated a few comments to help steer understanding of the
`$DEST_BRANCH` purpose.
Signed-off-by: Chris Evich <cevich@redhat.com>
Various tasks and scripts behave differently depending on whether or not
the build is running against a PR or on a branch, post-merge. However,
a great number of them are hard-coded to the string 'master' as the
destination. Since this is not always the case (there are other
relevant branches), it makes sense to abstract the references with a
single definition.
Add a top-level `$DEST_BRANCH` variable to CI, and otherwise
default to 'master' when unset. This enables running CI builds on
additional branches without the overhead of updating all the static
references to 'master'. Simply update `$DEST_BRANCH` at the top-level
and all branch-conditional logic will function as intended.
Signed-off-by: Chris Evich <cevich@redhat.com>
Over time unless they're removed, the project could grow quite a large
collection of VM images. While generally cheap (less than a penny each,
per month), these will become a significant cost item if not kept
in-check.
Add a specialized container for handling image-pruning, but limit
it to only finding and printing (not actually deleting) images.
Also update the image-building workflow so that base-images used to
compose cache-images are also labeled with metadata.
N/B: As an additional safeguard, the service account which
executes the new container in production *DOES NOT*
have access to delete images. This can be enabled
by adding the GCE IAM role: CustomComputeImagePrune
Signed-off-by: Chris Evich <cevich@redhat.com>
Normally when testing PRs, the final task to run is 'success'. It's
purpose is three-fold:
- Notify on IRC that a PR passed all testing.
- Block merging of a PR unless all dependent tasks are successful.
- When successful, publish cached binary release archives.
Mistakenly, the 'release' task was not made dependent upon the 'success'
task. Since 'success' only runs for PRs, this was causing post-merge
failures due to the 'release' task not finding any release archives -
the tasks which generate them are still running.
Fix this by making the 'release' task depend upon the same items as
the 'success' task. This will ensure it only runs as the very last
step, for both PRs and on branches (post-merge).
Signed-off-by: Chris Evich <cevich@redhat.com>
It's desirable to make archives available of builds containing actual
tested content. While not official distro-releases, these will enable
third-party testing, experimentation, and development for both branches
(e.g. "master") and pull requests (e.g. "pr3106").
* Add a Makefile targets for archiving both regular podman binaries
and the remote-client. Encode release metadata within these
archives so that their exact source can be identified.
* Fix bug with cross-compiling remote clients for the Windows and Darwin
platforms.
* Add unit-testing of cross-compiles for Windows and Darwin platforms.
* A few small CI-script typo-fixes
* Add a script which operates in two modes:
1. Call Makefile targets which produce release archives.
Upload the archive to Cirrus-CI's built-in caching system
using reproducible cache keys.
2. Utilize reproduced cache keys to attempt download of cache
from each tasks. When successful, parse the file's
release metadata, using it to name the archive file. Upload
all recovered archives to a publicly accessible storage bucket
for future reference.
* Update the main testing task to call the script in mode #1 for
all primary platforms.
* Add a new `$SPECIALMODE` task to call the script in mode #1 for
Windows and Darwin targets.
* Add a new 'release' task to the CI system, dependent upon all other
tasks. This new tasks executes the script in mode #2.
* Update CI documentation
Signed-off-by: Chris Evich <cevich@redhat.com>
Building/installing dependencies from fixed source-version ensures
testing is reliable, but introduces a maintenance burden and
risks testing far outside of a real-world environment. The
sensible alternative is to install dependencies from distro-packaging
systems.
Install all development and testing dependencies at VM cache-image build
time, to help ensure testing remains stable. The existing cache-image
build workflow can be utilized at any future time to build/test
with updated packages.
***N/B***: This does not update any dockerfiles used by testing, that is
left up to future efforts.
Signed-off-by: Chris Evich <cevich@redhat.com>
I'm running the BATS tests manually once in a while, and
catching several problems each week that make it past
the rest of CI. Since the BATS tests run at RPM gating
time, we need to catch problems earlier. Try running
the tests from Cirrus.
Tests will be skipped on Ubuntu due to a too-ancient
version of coreutils (8.28; the 'timeout -v' we use
requires 8.29).
Tests are run *after* integration tests, even though
these take three minutes and would be nice to have
fail quickly, because running before causes bizarre
CI failures. Shrug.
UPDATE: also fix run test, broken by #3311.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Remove disused `build_cache_images` task, and
update relevant dockerfiles for F30.
Fix problem of cloud-init failing to expand root-device on boot
(/var/lib/cloud/instance left in improper state).
Fix problem of cloud-init racing with google-network-daemon.service on
boot (looking for cloudconfig metadata too early). Causing
root-device to _sometimes_ fail to expand.
Fix problem of hack/get_ci_vm.sh argument passing.
Signed-off-by: Chris Evich <cevich@redhat.com>
New base-image boots, a cache-image builds, but more work is needed for
it to be prime-time ready. This commit just adds some updates to the
scafolding necessary to build the base-image. Future work will make F30
more of a reality.
Also add log-collection scripts to test image verification task
Signed-off-by: Chris Evich <cevich@redhat.com>
Previously, it was quite difficult to affect changes to VM cache images
without lots of manual work. This commit adds a new optional testing
task which mirrors the official-image build task which only runs on
master. In contrast, the new task may be run at any time in a PR, but
including a magic phrase in the PR description.
Update documentation to describe the new task and inform on it's usage.
Signed-off-by: Chris Evich <cevich@redhat.com>
We want the remote tests for our distributions to be tested in a
different VM than the local tests. This allows for faster CI runs and
easier debug as well as seperation of flakes.
Signed-off-by: baude <bbaude@redhat.com>
build a podman-remote binary for windows that allows users to use the
remote client on windows and interact with podman on linux system.
Signed-off-by: baude <bbaude@redhat.com>
Also rename `master_script` -> `failed_master_script` to clarify it's
"place in the line" when viewing (Cirrus WebUI)
Signed-off-by: Chris Evich <cevich@redhat.com>
The build_each_commit task builds each commit in a pull request
to verify that we have a (at least minimally) functional Podman
at every point, to aid in bisecting.
This task is, right now, extremely slow, taking around 1m40s to
build each commit - which quickly grows unreasonable as PRs grow
to 10+ commits.
Upping resources available to the task should decrease time spent
in CI and reduce the risk of hitting timeouts.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Previously libpod CI was fairly straight-forward, run unit and
integration tests in a standard set of 3 VMs. Off on the side was a
single special case of running tests as an ordinary user. There is a
desire to stop using the PAPR system to support testing inside of a
container.
Since having two special cases potentially invites more
down the road, make provisions to handle them more gracefully. This
commit introduces an environment variable: ``$SPECIALMODE``. It's
value has the following meanings within the CI scripts:
Mode 'none': Nothing special, business as usual (default)
Mode 'rootless': Rootless testing
Mode 'in_podman': Build container, run integration tests in it.
This will make adding additional special-cases later easier, as well as
extending the special cases in a Matrix across multiple OS's.
Signed-off-by: Chris Evich <cevich@redhat.com>
This task is heavily network bound, and takes way too long to include
along with the gating task. Separate it out and run it in parallel
to everything else. Also add some reasonable timeouts.
Signed-off-by: Chris Evich <cevich@redhat.com>
* Randomize the user's UID and GID
* Simplify `setup_environment.sh`
* Support new "-r" option for `hack/get_ci_vm.sh` setting up rootless
* Connect as $ROOTLESS_USER when using "-r" with `hack/get_ci_vm.sh`
Signed-off-by: Chris Evich <cevich@redhat.com>
Since `on_failure` notification is working, and IRC activity
is picking up, these messages mostly just get in the way.
Leave the PR-testing success messages in place for now,
since they're helpful when someone's waiting.
Signed-off-by: Chris Evich <cevich@redhat.com>
Until recently it was very difficult to execute any scripts if part of a
task failed. A new feature in Cirrus-CI makes this easy. Use it to
post a notice on IRC when any task fails.
Also: Add quotes around yaml-string values for consistency and
syntax-highlighting correctness.
Signed-off-by: Chris Evich <cevich@redhat.com>
to protect against regressions, we need to add a few gating tasks:
* build with varlink
* build podman-remote
* build podman-remote-darwin
we already have a gating task for building without varlink
Signed-off-by: baude <bbaude@redhat.com>
Make use of the built imgts container image to track
VM image usage statistics for every automation run.
Also update and add small check to the gate test
that verifies expected formatting/content of the
`.cirrus.yml` file WRT VM Image names.
Signed-off-by: Chris Evich <cevich@redhat.com>
Certain integration tests require execution as a regular user.
This is acomplished by `PodmanTest.PodmanAsUserBase()` wrapping a
specialized execution environment, in `test/utils/utils.go`. However,
doing this requires passing through python, which vastly increases the
complexity of debugging low-level problems.
This commit introduces a new parallel task, run as a regular user on the
VM as set by three environment variables. All commands executed in the
``rootless_test.sh`` script, will occur as a real user with a name and
home directory, just as `$DIETY` intended. All env. vars established
during `environment_setup.sh` (for root) are available. The PR source
in `$GOSRC` and `$GOPATH` are owned by this user, and ready for use.
Signed-off-by: Chris Evich <cevich@redhat.com>
It was found that after updating the cache-image for RHEL 7.6,
unpredictable integration tests start failing with 'permission denied'
errors. Much effort was spent trying to diagnose this, as all other
platforms pass.
Since running the latest/greatest podman on RHEL, will never be
officially supported (by Red Hat - must use RPM podman) this matrix-item
is of questionable value. Therefor, this commit disables RHEL testing.
Signed-off-by: Chris Evich <cevich@redhat.com>
We had a regression on master where we broke the build for
non-Varlink builds. Catch this in CI in the future.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
There is no native package for this, so the packaged version must also
be installed, otherwise all the support/dependencies would be removed
also (like go-md2man). Fix this by installing from the google released
tarball, into /usr/local/go and set $GOROOT to point there.
Also, include a small fix for hack/get_ci_vm.sh not installing
testing dependencies because of an old assumption.
***CIRRUS: REBUILD IMAGES***
Signed-off-by: Chris Evich <cevich@redhat.com>
* Make sure that all vendored dependencies are in sync with the code and
the vendor.conf by running `make vendor` with a follow-up status check
of the git tree.
* Vendor ginkgo and gomega to include the test dependencies.
Signed-off-by: Chris Evic <cevich@redhat.com>
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
Add necessary tweaks to base-image build/import process so that image
can successfully boot with networking enabled. Build the base image and
update ``.cirrus.yml`` accordingly.
Also make a minor improvement to cache-image building to help save space in
GCP image storage. Namely, instead of storing every produced image, export
them into google storage buckets w/ life-cycle enabled. For production use,
they may be converted back into GCE images, otherwise they will expire after
a time.
Signed-off-by: Chris Evich <cevich@redhat.com>
Once built, this container can be utilized by automation to help keep
track of VM images. All parameters are passed in via env. vars.
Signed-off-by: Chris Evich <cevich@redhat.com>
Previously it was not possible to specify keys from the ``env`` section
in the various GCE sections. Now that features is added, consolidate
all the cache image definitions into a single place, reducing
maintenance burden.
This also results in the names passing through into the VMs. This is
useful, e.g. for future tracking of image usage statistics.
Update get_ci_vm script hints for new image name definition format
Signed-off-by: Chris Evich <cevich@redhat.com>
The limit of build_each_commit seems to be 17 commits - any more
and it times out. Give it a bit more time to work with.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Fixes:
git rebase origin/ -x make
fatal: Needed a single revision
invalid upstream 'origin/'
make: *** [Makefile:351: build-all-new-commits] Error 1
By not running this test post-merge.
Signed-off-by: Chris Evich <cevich@redhat.com>
The podbot messages are becoming obnoxious as more distributions are
tested. Only call the `success.sh` script once, after all testing was
successful. Also make update the message to include more helpful text
and url.
Signed-off-by: Chris Evich <cevich@redhat.com>
Frequently debugging of CI-related problems requires going hands-on
within the environment. However, reproducing the environment by hand is
very tedious and error prone. This script permits authorized users to
produce VM's based on any available cache-image, and automatically remove
them upon logout.
Also: Bump up VM disk sizes to 200GB due to performance reasons
Signed-off-by: Chris Evich <cevich@redhat.com>
Since the most recent TWO versions of Fedora are officially supported
upstream, both need to be tested. Implement the concept of a 'prior'
Fedora release in both base-image and cache-image production. Utilize
the produced cache-image to test libpod. Remove F28 testing from PAPR.
Much thanks to @baude @giuseppe for help with this.
Signed-off-by: Chris Evich <cevich@redhat.com>
