Handling Missing Dependencies
When you export data from FlowForce Server, you can always exclude certain objects from the export. However, some objects may have dependencies on other objects. If you do not export dependencies together with the object that depends on them, this may lead to errors when you later import that data back into FlowForce Server. The example below is meant to help you understand the implications and how to address them.
Let's assume that you have the two records shown below and choose to export from FlowForce Server only the job, without exporting the credential record:
In this case, the following happens:
1.The job will be exported (but it will have a missing dependency)
2.The credential record will not be exported.
If you later attempt to import the .zip archive into a FlowForce Server environment where the object /public/Jobs/my.credentials exists, the import is possible because the missing dependency is now resolvable.
However, if the referenced credential does not exist in the target environment, the import dialog box looks as follows:
In the image above, notice that the credential record is no longer recognized as such—namely, its type is "configuration" instead of "credential", and the icon is different. The type "configuration" indicates a generic configuration object (another job or credential), whose type is not known. In this example, the "my.credentials" configuration object was not exported, and the exported package has no information about its exact type*, other than the reference path. Therefore, attempting to import the data above into FlowForce Server will result in an error like: "Operation failed: Path does not exist".
To fix this error, create the missing record at the path indicated by the error message (in this case, the "my.credentials" record), and then perform the import again.
* A credential reference may be a reference to a standalone credential object, and, in some cases, to a job which contains local credentials, see also Referring to Credentials from Jobs.