import-data-mappings

title	sidebar_position
Import Data Mappings	2

Import Data Mappings

There are situations where the values in the DataObjects are structured differently than the import parameters of the blueprint, so that a mapping is necessary. The present example demonstrates such an import data mapping. It is part of the Installation.

An import data mapping can for example be useful in the following situation. Suppose there are two Installations, and the first one exports DataObjects which the second one imports. If then the types of the export and import parameters do not match exactly, an import data mapping in the second Installation can adapt the data.

One can also use an import data mapping to set a fixed value for an input parameter, so that it is not necessary to create an extra DataObjects for the value.

The Example Mapping

We use the same blueprint as in the Import Parameters example. Recall that it has a string parameter text, and another parameter release which is an object with two string fields name and namespace.

On the other hand, let's consider the DataObjects. Suppose the release name and namespace are given in two separate DataObjects dataobject-name.yaml and dataobject-namespace.yaml. In the imports section of the installation, we load them into variables temp-name and temp-namespace:

imports:
  data:
    - name: temp-name                 # temporary variable
      dataRef: my-release-name        # DataObject
    - name: temp-namespace            # temporary variable
      dataRef: my-release-namespace   # DataObject

Now we use these variables in the import data mapping. The import data mapping is a template that creates the data that the blueprint expects, i.e. a release object and a text string. We build the release object from the two variables, and we set the text parameter to the constant value hello:

importDataMappings:
  release:                            # import parameter of the blueprint
    name: (( temp-name ))
    namespace: (( temp-namespace ))
  text: hello

Note that you must use spiff, rather than GoTemplate as templating language in the import data mapping. The reason is that the import data mapping belongs to the yaml manifest of the Installation, and a GoTemplate would in general not be well-formed yaml.

For more details, see Import DataMappings

The Default Mapping

It is possible to skip the import data mapping for some or all import parameters of a blueprint, if no mapping is necessary, i.e. if the DataObjects contain the data already in the format that the blueprint expects. Actually, this is what happened in the Import Parameters example. There we have loaded the DataObjects into variables release and text, whose names and structures matched already with the parameters of the blueprint:

imports:
  data:
    - name: release
      dataRef: ...
    - name: text
      dataRef: ...

As there was no import data mapping defined, by default the variables were mapped unchanged to the corresponding import parameters of the blueprint, i.e. they are mapped as if there would have been the following trivial import data mapping:

importDataMappings:
  release: (( release))
  text: (( text ))

Procedure

The procedure is as follows:

1. Add the kubeconfig of your target cluster to your target.yaml.

2. On the Landscaper resource cluster, create namespace example and apply the context.yaml, the dataobject-name.yaml, the dataobject-namespace.yaml, the target.yaml, and the installation.yaml:

   kubectl create ns example
   kubectl apply -f <path to context.yaml>
   kubectl apply -f <path to dataobject-name.yaml>
   kubectl apply -f <path to dataobject-namespace.yaml>
   kubectl apply -f <path to target.yaml>
   kubectl apply -f <path to installation.yaml>

3. To try out the echo server, first define a port forwarding on the target cluster:

   kubectl port-forward -n example service/echo 8080:80

   Then open localhost:8080 in a browser.

   The response should be "hello", which is the text defined in the import data mapping in the installation.yaml.

References

Import DataMappings

spiff