Do not expose the content of the manifest statically and instead rely on
the public methods to retrieve source specifications dynamically.
Since the methods require iterating through the pipelines to collect
source specifications, we should avoid calling the function multiple
times when we can reuse the returned values.
The new test_distro's manifest produces a slightly different empty
manifest when serialized even without content. Cloud API and Koji tests
have been adapted to match.
Weldr tests have been updated in several ways:
- The test_distro content resolver is used to resolve manifest content
before serializing.
- The test scenarios in TestCompose have been named for easier
troubleshooting (easier to identify a failing test by name).
- Manifests that work with the secondary ostree repo (the "other") use
the appropriate URL and ref and create a secondary "other" serialized
manifest.
The weldr API's test flag for resolving ostree commits does not produce
the same, fixed hash every time but instead computes a sha256 from the
URL + ref, like we do in the test manifests.
Initialise the manifest only once in the enqueue functions
(ImageType.Manifest()) and pass it to the manifest generation function
with the workers and the dependency IDs. The function is renamed from
generateManifest() to serializeManifest() to reflect its new
functionality more accurately. The arguments to the function have also
been trimmed since we no longer need the image type, blueprint, and
image options.
The new functionality of the function is to collect all the resolved
content from the workers and serialize the manifest object.
Use the container sources provided by the content on the initialised
manifest instead of the blueprint to resolve containers. The container
sources on the manifest are a map keyed by the name of the pipeline that
will use the resolved containers, but the worker's container resolve job
works on a slice, so we reread the content map to get the pipeline name
(instead of taking the first payload pipeline from the image type).
This requires that there be only one pipeline that embeds containers,
which currently true of all our image types.
Use the commit sources provided by the content on the initialised
manifest instead of the image options to resolve commits. The ostree
sources on the manifest are a map keyed by the name of the pipeline that
will use the resolved commit spec, but unlike with the package sets, the
worker's commit resolve job works on a slice, so we reread the content
map to get the pipeline name. This requires that there be only one
pipeline that requires a resolved ostree commit, which is currently true
of all our image types.
Setting the default ostree ref on the image options before calling
Manifest() isn't needed anymore.
The FetchChecksum on ostree.ImageOptions was the resolved commit ID of
the parent ref to be pulled (for ostree commits and containers) or the
commit ID of the content ref (for ostree installers and raw images).
With the new process of manifest creation and serialisation, using the
image options to transport resolved content references is bad and
confusing. Image options should only reflect user and image type
options before any references are resolved. With this change, the
ostree.ImageOptions should only reflect the ostree-related options
specified by the user. Commit IDs will only be available after the
manifest is initialised when the commit sources are resolved (before
serialisation).
Before, this was done in the PackageSets() function.
The reason for this is that having an ostree ref affects package
selection (for example, it adds rpm-ostree). At the package selection
phase, it doesn't matter what the ostree ref is; it is just used to
determine if a pipeline is for an ostree-based image type and it doesn't
affect non-ostree-based image types because the image functions ignore
it.
This is only needed in the cloudapi now because other places have
switched to using the new order of operations, where the manifest is
generated after the ostree commit is resolved, so it's always added when
needed.
To add the container specs to Serialize(), we need to map them to the
payload (OS) pipeline. We assume the first name in the image type's
PayloadPipelines() list is the OS pipeline, which is true of all image
types right now but might not be necessarily in the future.
This is a temporary workaround. Eventually, the mapping will be set by
the image type itself when we use the container source specs attached to
the Manifest object.
Still not using the new process for generating the manifest exactly.
This commit only replaces the call to PackageSets() with a call to
Manifest() to get the Content.PackageSets. This is essentially the same
as before, when we were initialising the manifest object twice.
The Manifest() function does a tiny bit more work than PackageSets(),
but it's minimal and we gain the benefit of only having a single code
path and, although it's run twice, it's always run in the same way.
Pass the entire Blueprint to Manifest() instead of just the
Customizations. The goal is to combine the functionality of the
ImageType.PackageSets() and ImageType.Manifest() methods into one call.
Return manifest.Manifest from the Manifest() function without
serializing. The caller then has to call the manifest.Serialize()
function using the depsolved packages.
This moves towards changing the order of actions required to generate a
manifest. With this change, the manifest creation and depsolving can be
done independently, but this still requires instantiating the manifest
object twice (InstantiateManifest() is called in PackageSets() and
Manifest()), which we don't want to have to do.
Move the FactsImageOptions from distro to the new rhsm/facts package.
At the same time define the values we use as an enum, including the
"test-manifest" value.
Though the values don't really matter, the test value is defined first
so it takes the 0 value, which feels nicer conceptually.
The field in the distro.ImageOptions is changed to be a pointer to allow
for nil values.
Same as with the container SourceSpec, the struct specifies the required
information to resolve an ostree commit from a source (URL, ref, and
optional parent).
Renaming for consistency.
Move the subscription options from distro to its own package.
Now we can import the manifest package into the distro package (instead
of the other way around) so we can work with the manifest.Manifest type
in distro.
Explicitly specify the AMI boot mode in AWS upload target in Cloud API
compose handler. The value is determined based on image type's boot
mode.
Signed-off-by: Tomáš Hozza <thozza@redhat.com>
Convert some of the fields in the `RepoConfig` struct
to pointers. Since `RepoConfig` will be used to convert
custom repositories to an array of `osbuild.YumRepository`,
we need to ensure that fields that are not set explicitly
are not saved to the `/etc/yum.repos.d` repository files.
Update the internal RepoConfig object to
accept a slice of baseurls rather than a
single field. This change was needed to
align RepoConfig with the dnf spec [1].
Additionally, this change adds custom json
marshal and unmarshal functions to ensure
backwards compatibility with older workers.
Add json tags to the internal rpmmd config
since this is serialized in dnfjson.
Add unit tests to check the serialization
is okay.
[1] See dnf.config
Add the ListDigest to the container Spec struct and all its copies so we
can store list digests when they are available and pass them on to the
appropriate osbuild stages, sources, and inputs.
Copy the value whenever a spec is moved to a different representation.
This changes the `Manifest` function of the `ImageType`
interface so that any warnings detected during the
`checkOptions` step of the manifest initialization can
be propagated back to the Weldr-API (see next commit).
Signed-off-by: Irene Diez <idiez@redhat.com>
Expose the Directory and File customizations in the Cloud API. Also
validate the provided customizations while processing the compose
request by trying to convert them to internal representations
`fsnode.File` and `fsnode.Directory`.
Signed-off-by: Tomáš Hozza <thozza@redhat.com>
Add new error `ErrorInvalidCustomization` used in situations when the
user-provided customization values don't pass validation. This will be
used by the Directory / File customization.
Signed-off-by: Tomáš Hozza <thozza@redhat.com>
When rhc is selected it will install the required packages, register
using rhc and always enable insights.
When rhc is not selected it will use subscription manager for
registration, and optionally enable insights. Also installing required
packages.
DNF supports more than one GPG key. It is possible that one may be used for
signing packages, and another to sign the repository metadata. This
renamed GPGKey to GPGKeys internally. It does not change the on-disk
repository json format.
We have switched how 5xx errors are being recorded
internally and we are now recording all failures
for all endpoints. As a result, a dedicated metric
only for compose failures is no longer required.
After introducing Go 1.18 to a project, it's required by law to convert at
least one method to a generic one.
Everyone hates IntToPtr, StringToPtr, BoolToPtr and Uint64ToPtr, so let's
convert them to the ultimate generic ToPtr one.
Signed-off-by: Ondřej Budai <ondrej@budai.cz>
The ostree options are used during the ostree resolve job, but when
generating the manifest the rhsm value comes from the image options, so
it's necessary to set it on both.
Register the custom middleware function to the cloudapi
server. This function is responsible for recording all
the status codes of all the server's endpoints.
Specifically describe that at least one of `baseurl`, `mirrorlist` or
`metalink` properties must be specified. Although this is not enforced
by the schema, it is the reality. Modifying the schema to enforce this
results in the generated data structure for `Repository` to be simply
an `interface{}`, which would just make the code more ugly, Therefore
this is covered by the API implementation logic.
Signed-off-by: Tomáš Hozza <thozza@redhat.com>
Providing the `location` is no longer required for Azure Upload Options.
If it is not provided, the implementation determines the location from
the provided Resource Group. This will make the API nicer for any
client, since they won't need to provide redundant information.
Signed-off-by: Tomáš Hozza <thozza@redhat.com>
Koji clean-up doesn't delete the top-level directories. As we named every
top-level directory different, they were just cumulating on kojihub, which
is obviously wrong.
This commit changes that behaviour to put all the temporary directories under
a new osbuild-cg top-level one. This way, osbuild-cg won't ve ever cleaned,
whereas osbuild-cg/osbuild-composer-koji-{UUID} will be, which is exactly
what we want.
Closes: https://github.com/osbuild/osbuild-composer/issues/3064
Signed-off-by: Ondřej Budai <ondrej@budai.cz>
