All soluble entities in the graph have a URN. We found Amazon's ARN naming scheme to be extremely useful and effective for maintaining logical references in the knowledge graph. They are globally unique and have semantic value.
Comparison to UUID
Systems like Kubernetes have unique identifiers (uid) that are for all practical purposes globally unique. These synthetic identifiers are very effective for establishing uniqueness. Unfortunately, when building a knowledge graph, there are many cases where this identity uniquess is not desirable.
To use Kubernetes as an example if we have a cluster
mycluster, with a service named
auth in the
default namespace, it will have a globally unique
uid. That is great. However, if we delete that service and recreate the service in the same cluster with the same name and namespace, it will have a new
uid. The problem is that for most practical purposes, the new and the old service is the same thing. The fact that the
uid is different is largely an implementation detail that is irrelevant for any practical application.
Most Systems Don't Have URN
Most systems don't have native URN as Amazon does but they do have all the necessary attributes. GCP has something similar with its REST URLs for resources, which are easily adapted to a URN format. Others have to be created based on our understanding of the target system.
By default form login is enabled. A user named
admin will be seeded with a password of
The dashboard supports SAML authentication.
To configure SAML, you will need to obtain a SAML metadata XML file from your identity provider. You should either make that file available at
SAMLE_METADATA to the location of the file.
SAML_RESPONSE_URL are also required for most identity providers. Okta is the notable exception.
|SAML_METADATA||Path to SAML metadata XML file. If not specified, |
|SAML_ENTITY_ID||Entity ID of soluble dashboard, as configured in your SAML IDP.|
|SAML_RESPONSE_URL||Assertion response (ACS) URL. This will be |