Trusted Supply Chain Component Registry (TSCCR) enforcement starts Monday and an
internal report shows our semgrep action is pinned to a version that's not
currently permitted. Update all the action versions to whatever's the new
hotness to maximum the time-to-live on these until we have automated pinning
setup.
Also version bumps our chromedriver action, which randomly broke upstream today.
this is basically to avoid Fear/Uncertainty/Doubt
the github action actions/setup-go
(and, with a different chache key, hashicorp/setup-golang)
caches both GOMODCACHE (go source files), which is good,
and GOCACHE (build outputs), which *might* be bad,
if the cache was built on an OS with an older glibc
than we want to support. from `go help cache`:
> [...] the build cache does not detect changes to
> C libraries imported with cgo.
so in enterprise we can use Vault for secrets,
without merge conflicts from oss->ent.
also:
* use hashicorp/setup-golang
* setup-js for self-hosted runners
they don't come with yarn, nor chrome,
and might not always match node version.
Update the revision used by the docker action. This should always reflect the commit that's being built as this may differ from the default <github.sha> that the workflow was invoked at.
Goes with https://github.com/hashicorp/actions-docker-build/pull/59 - and should not be merged until this PR is merged and a new version of the action is cut.
Although #17669 fixed the permissions of the release pipeline to push new
commits, there was still an error when invoking the `build` workflow.
The format of the reference was changed in #17103 such that we're sending the
git ref (a SHA) and not the "--ref" argument required by the GH actions workflow
API, which in this case is apparently specially defined as "The branch or tag
name which contains the version of the workflow file you'd like to run" and not
what git calls a "ref".
This changeset:
* Removes the third-party action entirely so that we're using GitHub's own
tooling. This removes one more thing from the supply chain to pin and ensures a
1:1 mapping of args to what's documented by GitHub.
* Removes the `--ref` argument entirely, which causes it to default to the
current branch that the release workflow is running on (which is always what
we want).
Since the matrix exercises different test cases, it's better to allow
all partitions to completely run, even if one of them fails, so it's
easier to catch multiple test failures.
In #17103 we set read-only permissions on all the workflows. Unfortunately we
missed that the `release` workflow makes git commits and pushes them to the
repository, so it needs to have write permissions.
Some of the paths ignored by `test-core.yaml` need to be checked by
`make check`. The `checks.yaml` workflow run on these paths and can also
be used as a reusable workflow.
This seems to finish in about 20 minutes, or run for 6+ hours until hitting
a default timeout. Set a timeout to 30 minutes so we aren't wasting time
and runners.
* Exam to parallelize tests
* Logging to try to solve test flakiness
* Logging in another failure
* Hardening for one test and snapshot for another
* Explicitly set the first one as the servicedAlloc instead of randomly picking
* A wild CircleCI test failure appears
* de-log
namely, these workflows:
test-e2e, test-ui, and test-windows
extra-curricularly, as part of the overall
migration effort company-wide, this also includes
some standardization such as:
* explicit permissions:read on various workflows
* pinned action version shas (per https://github.com/hashicorp/security-public-tsccr)
* actionlint, which among other things runs
shellcheck on GHA run steps
Co-authored-by: emilymianeil <eneil@hashicorp.com>
Co-authored-by: Daniel Kimsey <daniel.kimsey@hashicorp.com>
The 32-bit Intel builds (aka "386") are not tested and likely have bugs
involving platform-sized integers when operated at any non-trivial scale. Remove
these builds from the upcoming Nomad 1.6.0 and provide recommendations in the
upgrade notes for those users who might have hobbyist boards running 32-bit
ARM (this will primarily be the RaspberryPi Zero or older spins of the RaspPi).
DO NOT BACKPORT TO 1.5.x OR EARLIER!
The file path in the TSCCR repo for the `returntocorp/semgrep` action was
incorrect, so the pinning tool was not able to find the correct entry and it was
not pinned in #17238.
The repository is fixed in https://github.com/hashicorp/security-tsccr/pull/431
The `backspace/ember-asset-size` action we're using is unmaintained and has a
bunch of vulns in it, so it won't pass security screening (this is a NodeJS
action so it has piles of dependencies, 99% of which won't be in use but fails
automated screening anyways). Move this to the upstream version.
The `machine-learning-apps/pr-comment` action also presents a problem for the
ProdSec security screening because it's archived and also runs an external
Docker image. Move this to a likely-ok maintained action for now, until we can
spare some time to remove this in lieu of something more reasonable that isn't a
GitHub action.
Many of the GitHub Actions from the build pipeline are written in a truly
ancient version of NodeJS. Upgrade to more recent versions.
Remove RelEng from codeowners
This reverts commit 1721e687c0832bea3d9b7eec5dcd3c4e7a924d71.
The change was expected to solve the sporadic problems we were having
with Backport Assistant, but it end up creating even more failures.
Examples in the documentation frequently include tokens, including Vault tokens
which end up triggering GitHub's secret scanner. Remove these from consideration
so that we don't get false positive reports.
Instead of attempting to pick each individual commit in a PR using
`BACKPORT_MERGE_COMMIT` only picks the commit that was merged into
`main`.
This reduces the amount of work done during a backport, generating
cleaner merges and avoiding potential issues on specific commits.
With this setting PRs that are not squashed will fail to backport and
must be handled manually, but those are considered exceptions.
Running tests `on: push` prevents GitHub from showing the workflow approval
button, which prevents tests from being run on community-contributed (or even
just non-Nomad HashiCorp folks) PRs. Running `on: pull_request` automatically
picks up opened, reopened, and synchronize hooks (where "synchronize" means a
push to HEAD on the PR's branch, so that'll pick up rebases and updates).
But we also want to run tests on `main` and the various `release` backport
branches, so retain a `on: push` for those.
This PR tries to make API tests run fast, as an experiment to later apply
to all packages. Key changes include
- Swapping freeport for test/portal for port allocations
- Swappng some uses of WaitForResult with test/wait
- Turning on parallelism in api/testutil/slow.go
- Switching to custom public runner (32 vcpu)
There's also chunk of cleanup brought in for the ride
