Git Repository Contents
SUSE® Rancher Prime Continuous Delivery will create bundles from a git repository. This happens either explicitly by specifying paths, or when a fleet.yaml is found.
The folder could contain a Helm chart, or reference one. It could be a plain Kubernetes manifest, or a Kustomize folder. Each bundle is converted to a single Helm chart for deployment.
The fleet.yaml file contains all the options for the deployment.
Bundle Names
By default, bundle names will also be generated from the GitRepo’s name and the path from which the bundle is created. However, a bundle’s name can be overridden by using the name field in a fleet.yaml file.
Bundle lifecycles are tracked between releases by the Helm releaseName field added to each bundle. If the releaseName is not specified within fleet.yaml it is generated from GitRepo.name + path. Long names are truncated and a -<hash> prefix is added.
How repos are scanned
The git repository has no explicitly required structure. It is important
to realize the scanned resources will be saved as a resource in Kubernetes so
you want to make sure the directories you are scanning in git do not contain
arbitrarily large resources. Right now there is a limitation that the resources
deployed must gzip to less than 1MB.
Multiple paths can be defined for a GitRepo and each path is scanned independently.
Internally each scanned path will become a bundle that SUSE® Rancher Prime Continuous Delivery will manage,
deploy, and monitor independently.
The following files are looked for to determine the how the resources will be deployed.
| File | Location | Meaning |
|---|---|---|
Chart.yaml |
/ relative to |
The resources will be deployed as a Helm chart. Refer to the |
kustomization.yaml |
/ relative to |
The resources will be deployed using Kustomize. Refer to the |
fleet.yaml |
Any subpath |
If any fleet.yaml is found a new bundle will be defined. This allows mixing charts, kustomize, and raw YAML in the same repo. |
.yaml |
Any subpath |
If a |
overlays/{name} |
/ relative to |
When deploying using raw YAML (not Kustomize or Helm) |
Alternative scan, explicitly defined by the user
In addition to the previously described method, SUSE® Rancher Prime Continuous Delivery also supports a more direct, user-driven approach for defining Bundles.
In this mode, SUSE® Rancher Prime Continuous Delivery will load all resources found within the specified base directory. It will only attempt to locate a fleet.yaml file at the root of that directory if an options file is not explicitly provided.
Unlike the traditional scanning method, this one is not recursive and does not attempt to find Bundle definitions other than those explicitly specified by the user.
Example File Structure
driven
|___helm
| |__ fleet.yaml
|
|___simple
| |__ configmap.yaml
| |__ service.yaml
|
|___kustomize
|__ base
| |__ kustomization.yaml
| |__ secret.yaml
|
|__ overlays
| |__ dev
| | |__ kustomization.yaml
| | |__ secret.yaml
| |__ prod
| | |__ kustomization.yaml
| | |__ secret.yaml
| |__ test
| |__ kustomization.yaml
| |__ secret.yaml
|__ dev.yaml
|__ prod.yaml
|__ test.yaml
Corresponding GitRepo Definition
kind: GitRepo
apiVersion: fleet.cattle.io/v1alpha1
metadata:
name: driven
namespace: fleet-local
spec:
repo: https://github.com/0xavi0/fleet-test-data
branch: driven-scan-example
bundles:
- base: driven/helm
- base: driven/simple
- base: driven/kustomize
options: dev.yaml
- base: driven/kustomize
options: test.yaml
In the example above, the user explicitly defines four Bundles to be generated.
-
In the first case, the base directory is specified as
driven/helm. As shown in the directory structure, this path contains afleet.yamlfile, which will be used to configure the Bundle. -
In the second case, the base directory is
driven/simple, which contains only Kubernetes resource manifests (configmap.yamlandservice.yaml). Since nofleet.yamlor options file is specified, SUSE® Rancher Prime Continuous Delivery will generate a Bundle using the default behavior—simply packaging all resources found within the directory. -
The third and fourth cases both reference the same base directory:
driven/kustomize. However, each specifies a different options file (dev.yamlandtest.yaml, respectively). These options files define overlay-specific configuration for each environment (e.g., dev, test) by selecting the appropriate kustomize overlay subdirectories and applying them on top of the shared base.
SUSE® Rancher Prime Continuous Delivery will process these as distinct Bundles, even though they originate from the same base path, because the provided options files point to different configurations.
An example of the files used in the third and fourth Bundles would be the following: (These files follow the exact same format as fleet.yaml, but since we can now reference them by name, we can use one that best suits our needs)
namespace: kustomize-dev
kustomize:
dir: "overlays/dev"
It is important to note that any path defined in these files must be relative to the base directory used when the Bundle was described.
For example, with the previously mentioned structure, we are defining the base directory as driven/kustomize. That is the directory we need to use as the root for the paths used in Kustomize files.
We could decide to place the dev.yaml file at the path driven/kustomize/overlays/dev (this is supported), and then define the Bundle as:
bundles:
- base: driven/kustomize
options: overlays/dev/dev.yaml
However, the path defined within dev.yaml should still be relative to driven/kustomize.
This is because when SUSE® Rancher Prime Continuous Delivery reads the options files, it always uses the base directory as the root.
In other words, with the previous example… this would be incorrect:
namespace: kustomize-dev
kustomize:
dir: "."
And the correct definition should still be:
namespace: kustomize-dev
kustomize:
dir: "overlays/dev"
With this new way of defining Bundles, SUSE® Rancher Prime Continuous Delivery becomes much more direct and also simplifies the adoption of deployments using kustomize. In the example, we can see a complete kustomize use case where for each Bundle, we can specify which version we want.
With the previous scanning option, SUSE® Rancher Prime Continuous Delivery cannot determine which YAML we want to use to configure the Bundle, so it attempts to find it on its own (Which, at times, does not provide enough flexibility.)
|
Excluding irrelevant files from user-scanned bundles When using this bundle scanning mode, SUSE® Rancher Prime Continuous Delivery does not exclude bundle configuration files which are not explicitly referenced in the GitRepo. For instance, in the above example file structure:
This can be mitigated by using a |
Excluding files and directories from bundles
SUSE® Rancher Prime Continuous Delivery supports file and directory exclusion by means of .fleetignore files, in a similar fashion to how .gitignore
files behave in git repositories:
-
Glob syntax is used to match files or directories, using Golang’s
filepath.Match -
Empty lines are skipped, and can therefore be used to improve readability
-
Characters like white spaces and
#can be escaped with a backslash -
Trailing spaces are ignored, unless escaped
-
Comments, ie lines starting with unescaped
#, are skipped -
A given line can match a file or a directory, even if no separator is provided
-
A match may be found at any level below the directory where a
.fleetignorelives -
Multiple
.fleetignorefiles are supported
root/
├── .fleetignore # contains `ignore-always.yaml'
├── something.yaml
├── bar
│ ├── .fleetignore # contains `something.yaml`
│ ├── ignore-always.yaml
│ ├── something2.yaml
│ └── something.yaml
└── foo
├── ignore-always.yaml
└── something.yaml
Unsupported:
-
Double asterisks (
**) -
Explicit inclusions with
!
fleet.yaml
The fleet.yaml is an optional file that can be included in the git repository to change the behavior of how
the resources are deployed and customized. The fleet.yaml is always at the root relative to the path of the GitRepo
and if a subdirectory is found with a fleet.yaml a new bundle is defined that will then be
configured differently from the parent bundle.
|
Helm chart dependencies: SUSE® Rancher Prime Continuous Delivery automatically handles updating Helm chart dependencies, unless flag If automatic dependencies updates are disabled, you must manually run:
Refer to Rancher’s SUSE® Rancher Prime Continuous Delivery docs for more information. |
The available fields are documented in the fleet.yaml reference. For a private Helm repo, users can reference a secret from the git repo resource. See Using Private Helm Repositories.
Using Helm Values
How changes are applied to `values.yaml`:
-
The most recently applied changes override previous values.
-
Merge order:
helm.values→helm.valuesFiles→helm.valuesFrom
The targeting step can treat values as templates using cluster info. More info: templates in fleet.yaml.
You can disable this using disablePreProcess.
Credentials in Values
If the chart generates credentials, override them or the chart may redeploy continuously.
Credentials loaded via valuesFrom are encrypted at rest if Kubernetes encryption at rest is enabled.
Using ValuesFrom
These examples showcase the style and format for using valuesFrom.
Propagating ConfigMaps and Secrets to downstream clusters: ConfigMaps and Secrets should generally be created directly in downstream clusters.
However, from Fleet v0.14.0 onwards, they can also be referenced through a HelmOp’s downstreamResources field to be
automatically propagated to targeted downstream clusters.
Refer to experimental downstream resources for more information.
Example ConfigMap:
apiVersion: v1
kind: ConfigMap
metadata:
name: configmap-values
namespace: default
data:
values.yaml: |-
replication: true
replicas: 2
serviceType: NodePort
Example Secret:
apiVersion: v1
kind: Secret
metadata:
name: secret-values
namespace: default
stringData:
values.yaml: |-
replication: true
replicas: 3
serviceType: NodePort
Referencing them:
helm:
chart: simple-chart
valuesFrom:
- secretKeyRef:
name: secret-values
namespace: default
key: values.yaml
- configMapKeyRef:
name: configmap-values
namespace: default
key: values.yaml
values:
replicas: "4"
Per Cluster Customization
The GitRepo defines which clusters a repo deploys to. The fleet.yaml defines customizations per target.
All clusters and groups in the same namespace are evaluated against targets. First match applies.
targetCustomizations:
- name: all
clusterSelector: {}
- name: none
clusterSelector: null
Matching by cluster name:
targetCustomizations:
- name: prod
clusterName: fleetname
Raw YAML Resource Customization
# Base files
deployment.yaml
svc.yaml
# Overlay files
overlays/custom/configmap.yaml
overlays/custom/svc.yaml
overlays/custom/deployment_patch.yaml
Rules:
-
Matching names replace base files.
-
_patch.files apply patches (JSON/strategic/JSONPatch).
Cluster and Bundle State
See status fields.