docs/static/v0.9/concepts/logical/registry/index.html
The Meshery Registry is a vital component within Meshery, serving as a centralized repository for managing a diverse range of cloud and cloud native resources. It stores and organizes crucial information such as models, categories, components, and relationships, enabling efficient interaction and utilization of these resources within the Meshery ecosystem. You can conveniently access and manage registry data through Meshery UI, and through Meshery CLI (mesheryctl registry).
As the central repository for all capabilities known to Meshery, the registry contains various entities.
#Contents of the Registry
Models: Blueprints defining configurations for interacting with cloud native infrastructure. They consist of operations, components, relationships, and policies.
You will find two types of models in the registry: Static and Dynamic.
Each Meshery release comes with a built-in set of models automatically registered at Meshery Server boot-time. These built-in models offer a core set of entities for Meshery’s supported integrations. Once Meshery Server is running, and as it connects to and discovers your infrastructure, dynamic models are automatically generated. A given Meshery release may not include all possible models found in your environment, so Meshery automatically generates and registers new models and components based on the specific infrastructure Meshery is connected to. Dyanmic models often lack additional metadata, such as descriptions, tags, and relationships, which are typically included in static models.
Use either Meshery UI or CLI to interact with the Registry. Meshery UI offers a user-friendly visual interface for browsing, searching, and managing registry entries. You can easily explore available models, components, and relationships, gaining insights into their properties and connections. Meshery CLI offers commands so that you can register, list, retrieve, update, and delete models, components, and relationships directly from the command line.
The process of generating a Model (and its entities) is a multi-step process and does not require use of Meshery Server. The process begins with the sourcing of the model information from an authoritative source: a Registrant. Registrants are responsible for providing all the necessary information to Meshery to generate the model.
Meshery CLI supports the generation of models from a Google Spreadsheet. The Google Spreadsheet should contain a list of model names and source locations from any supported Registrant (e.g. GitHub, Artifact Hub) repositories. The source locations can be a URL to a folder containing Kubernetes CRDs, or to a Helm Chart tar.gz, or an individual Kubernetes Manifest with custom resource definition.
See mesheryctl registry generate for more information.
Once registered in the Meshery Registry, Models and their entities are available for use within that specific Meshery Server.
Meshery Adapters are one example of a Registrant. Registrants are responsible for the registration of entities in the Meshery Registry. Adapters are responsible for the sourcing and registration of entities and the packaging of these enties into one or more models.
mesheryctl model import -f <path-to-model>
Visit the Settings Registry page and click the “Import” button to import a model.
You have control over whether a registered entity (model and all that the model contains) this can be an individual or team-level preference. Use the “Ignore” action to designate whether a given model is allowed to be used within a given Meshery Server deployment. Models that are ignored remain in the Meshery Registry but are not available for use within a given Meshery Server deployment.
[ Previous
Credentials ](/v0.9/concepts/logical/credentials)[ Next
Workspaces ](/v0.9/concepts/logical/workspaces)