Tenant resources¶
Tenant resources are the external systems a tenant can use: the GitHub repositories models are imported from, and the data locations used by import and export jobs. Select ‘Tenant resources’ from the ‘Manage tenant’ menu.
Resources are added and changed by an admin; other users can view them.
Model repositories¶
Model repositories are the GitHub repositories that models are imported from, listed on the ‘Model repositories’ tab. trac reaches them through your organisation’s GitHub App, so a platform admin must add the App’s credentials first. Until then, ‘Add GitHub repository’ is unavailable.
Note
In Community, this page is ‘Manage repositories’ on the ‘Configuration’ menu, and trac’s examples repository is already available. Instead of a GitHub App, you sign in to GitHub on this page using trac’s GitHub OAuth app or a personal access token, and that sign-in is used to reach private repositories.
Adding a repository¶
Click ‘Add GitHub repository’. Give the repository a Key, which is how it’s shown when people import a model, then fill in its properties:
Owner Type — whether the repository belongs to an organisation or a user
Owner — the GitHub organisation or user that owns it
Repository name
HTTP URL and API URL — the GitHub addresses trac connects to. These default to
https://github.comandhttps://api.github.com, and only need changing for GitHub Enterprise ServerRepository URL — the repository’s own URL
The GitHub App must be installed on the repository, or on all of the owner’s repositories, for trac to reach it.
Click ‘Save’ and confirm. The repository is listed, and can be selected when importing a model.
Note
The repository’s properties are checked when it’s saved, but the connection isn’t tested. Access problems, such as the GitHub App not being installed on the repository, show up when a model is first imported from it.
Editing a repository¶
Click ‘Edit’ at the bottom of a repository to unlock it, change its properties and save it again. The Key can’t be changed.
Deleting a repository¶
If a repository is deleted, trac can no longer use it to import models or run jobs. Models already imported will still appear in the platform but will no longer be executable.
Data locations¶
Not available in Community
Data locations are the external storage locations that data import and export jobs read from and write to. They’re listed on the ‘Data locations’ tab.
Click ‘Add data location’ and choose the type of storage. Only the types a platform admin has enabled are offered. Give the location a Key, which is how it’s shown when people set up import and export jobs, then fill in its properties:
Local storage |
The Root path of a folder on the server’s file system |
S3 |
The Bucket, and optionally a Prefix within it, a Region and an Endpoint. Credentials are either ‘Default’, using the platform’s own AWS credentials, or ‘Static’, with an Access key ID and Secret access key |
Azure |
The Storage account and Container, and optionally a Prefix. Credentials are ‘Default’ (the platform’s managed identity), ‘Account key’ or ‘SAS token’ |
Only the credential methods a platform admin has enabled can be chosen. Keys and tokens are stored securely and are never shown again once saved.
Every location also has a Storage access setting, marked ‘Platform’ because trac applies it rather than the storage itself:
Import — the location is offered only for import jobs
Export — the location is offered only for export jobs
Both — the location is offered for both
Save the location. It’s now available on the ‘Import data’ and ‘Export data’ pages.
Note
If a platform admin later disables a storage type or credential method, locations that use it stay listed but are marked unavailable, and jobs that use them fail before they run.
Documentation v0.11