Project Harbor has released version 1.6.0 with several new features such as Helm charts management, improved LDAP support, image replication, and database migrations. A big thank you to the community for your incredible continued support!
Let’s dive into it what’s new:
Helm has become the de facto package manager of Kubernetes, making it easy to to deploy a vast array of applications. We see a dedicated Helm chart repository as a must-have system to help build new enterprise IT infrastructure and platforms. Helm charts should work seamlessly together with container images, and therefore supporting both image management and Helm chart management has become the natural direction for Harbor.
Starting with version 1.6.0, Harbor can serve as a unified cloud native registry for both image management requests and helm chart management requests. The main features are grouped into two perspectives:
It’s easy to get started with Helm charts and Harbor using the Helm CLI:
helm repo addto add Harbor as a unified chart repository with specified username, all the project namespaces accessible by that user should be visible to Helm.
helm repo addto add a Harbor project as a separate chart repository, only the charts under that project are visible to Helm.
helmCLI to push charts to Harbor
helm installto download the chart from Harbor and install it to the target Kubernetes environment
helm verifyare also supported
For more details on managing Helm charts with Harbor, see the user guide.
Many Harbor administrators are using LDAP to authenticate Harbor users, and then need to assign roles to each user. To make this more scalable Harbor now supports assigning roles to LDAP groups as well as individual users.
For more details on managing roles with LDAP groups, see the docs.
Two replication filters (repository name filter and repository tag filter) were introduced in the previous version of Harbor, they both need the filter pattern to match the images which should be replicated. Sometimes this is difficult or even impossible. With the introduction of label filters, users can add any label to the images without changing the name of them.
Users replicate images by adding labels to them and creating a new rule with the label filter.
For more details on this, see the user guide.
In the previous releases, there are two or three database instances running on one Harbor node, which are MariaDB/MySQL and PostgreSQL. This of course increases the effort to maintain the Harbor system. This new feature makes it possible to merge multiple databases into a single database, making it far easier to maintain and make it possible to enable HA solutions for the future releases.
No matching community articles found at this time.