| Infrastructure |
- Edge offers more flexibility on where you can install your Edge site:
- For simple and straightforward installations, install on a bundled k3s cluster.
- For more control over your resources, install on a managed Kubernetes cluster:
- AKS
- AWS Fargate using EKS
- EKS
- GKE
- OpenShift
- You can reuse your existing virtual machine for an Edge site if it meets the system requirements.
- For an easier and more straightforward Edge experience, you can request aCollibra Cloud site.
- You can install edge on a dedicated or shared cluster.
- There are multiple methods to install Edge site on a managed Kubernetes cluster:
- The Edge CLI via the default or restrictive method, for more control.
- Helm chart
- For Collibra Self-Hosted environments, Edge offers 2 options for enhanced security and control:
- Edge for Collibra Self-Hosted
- Airgapped Edge
|
- You can only install Jobserver on a virtual machine.
|
| Capabilities |
- Edge offers a wider range of data source connections:
- Connection and capabilities are continuously updated on a regular basis.
- For self-hosted airgapped deployments:
- IPv6 compatible connections
- FIPS compliant capabilities
|
- Jobserver-certified JDBC connectors
- JDBC metadata ingestion: Schema level
- Data profiling
- Data classification
- Data sampling
- Some integrations
|
| Security |
- Edge provides transparency and security with images.
- Mirror images in your private docker registry to scan containers according to your security acceptance criteria and policies. This allows security stakeholders use this information to evaluate vulnerabilities in Edge containers and determine risk, tolerance, and remediation requirements.
- Edge delivers weekly releases, addressing security patches and vulnerabilities within Collibra's SLA.
- Integrate your Edge site with your existing vault provider to implement your organization’s credential management policies for any data source to which Edge connects.
- Data source secrets / credentials are stored on the Edge site.
- Classification is local to the Edge site, which means that none of your data leaves your premises.
- Edge requests sample data from the data source and caches it for 24 to 48 hours.
|
- Data source secrets / credentials are stored in the Collibra Cloud.
- To classify data, client sample data is sent to the Cloud Classification platform.
- Sample data is stored in the Collibra Cloud.
|
| Performance |
|
- All capabilities executed sequentially.
- Limit on table size you can profile.
|
| Extensibility |
Edge is a run time platform to host all capabilities, and this includes any new capabilities that Collibra develops in the future. |
Jobserver requires several separate components for new capabilities, such as Separate Jobserver for Tableau ingestion or separate command line application for lineage harvester. |
| Maintenance |
- Installer bundled with minimal binaries and artifacts. The installation is “live” in that the installer pulls images from the repository over the internet at install time.
- Edge provides 3 upgrade modes. Edge can be upgraded as soon as there is a release available. It can also be updated when the customer prefers and is typically to review security vulnerabilities in a release, a feature in Smart Upgrade.
- Automatic: Your Edge site is upgraded as soon as a new release version is available.
- Manual: You control when your Edge site is upgraded, allowing you to review security vulnerabilities in a release version before upgrading your Edge site.
|
- Console application on the Jobserver offers a user interface for some configuration and access to logs. Some configuration changes must be made directly in configuration files.
- Jobserver only has one upgrade mode: Manual.
|