Harbor Reviews

We use Harbor for Docker containers to build from our GitLab CI/CD pipeline. This pipeline then automatically pushes to our Harbor instance. We are still working on storing Helm charts in the Harbor instance as well.

Harbor has improved our organization's security and stability in our container management processes by providing comprehensive security and stability features.

The most valuable feature of Harbor is its security scanning capability. Harbor has a built-in Trivy security scanner that detects vulnerabilities or CVEs. Additionally, the private registry, proxying, and caching from public registries are crucial.

There is a need for more support for public repositories, like the Google Cloud GTR, which is not as integrated as desired. Additionally, more example implementations for installation using Ansible would be beneficial.

We have been using Harbor for four months, building our new Kubernetes platforms.

The product has been stable since its implementation.

Before Harbor, we worked with Nexus. Nexus supports a broader range of repositories for application development and operating system images, which Harbor does not offer. However, Nexus doesn't have good support for duplicate Docker images.

The initial setup was quite straightforward, although we are still configuring it with Ansible and OPNs for automation, which is time-consuming.

Return on investment has not been realized yet. It is probably too soon.

Harbor is open source and part of the graduated Cloud Native Computing Foundation (CNCF) landscape, making it a cost-effective solution.

Compared to Nexus and JFrog, JFrog may be better for enterprises that develop applications, supporting operating system patching, however, it is significantly more expensive at about $45,000 a year.

I would recommend Harbor to others currently. It is necessary to consider the specific use case, as JFrog can be more advantageous if you develop your own applications.

I'd rate the solution nine out of ten.

This is a cloud-native product for document and image storage.  

The security vulnerability detection is the most valuable feature. It has also the Registryctl (CTL stands for 'control') feature, so we can store our Docker images.  

Right now, we are not looking for many improvements in the product. We are still exploring it. But I think the GUI could be improved and there could be better storage options.  

We have been using it for the last two or three months. 

We have not had any issues with stability to this point.  

Right now we have not had to do anything with scalability because we are new to the product.  

Approximately 10 to 15 people are using it in our company.  

We have never used the technical support, so we never raised a ticket that we have been assisted with. I have no idea about that in firsthand experience.  

Similar products we have used include Nexus (Sonatype) and JFrog (Artifactory). In the public cloud, we have also been using AWS ECR (Amazon Web Services, Elastic Container Registry) as a data registry. For vulnerability scanning, there is one more open-source tool that we have used called Anchore.  

We decided to switch to using Harbor because there is nothing like it amongst these tools. For Nexus and JFrog we needed some vulnerability reports that they did not provide and we are getting these reports freely from Harbor. This Harbor tool also has the advantage of being an open-source tool and costs nothing. That is why we switched.  

The initial setup is straightforward. It took only two to three hours, maybe. What takes some time is the product has to check all of our documents for vulnerabilities.  

The product is opensource so the cost of the product is free for everyone.  

We evaluated other products that we used concurrently. That was Nexus, JFrog, AWS ECR, and Anchore.  

My advice to people looking into Harbor is that I think it is a good tool. There is no reason not to try and use it because it is a free solution for everyone. Everyone can give a try for that price.  

On a scale from one to ten (where one is the worst and ten is the best), I would rate Harbor as a seven-out-of-ten.  

The GUI interface can be better and more user-friendly. There may also be some opportunity to offer different options for storage. If we are working on AWS, we can store its data in the S3 (Simple Cloud Storage), but that is not very straightforward and complicates the processing.   

The storage means we mainly use right now is the EC2 (Elastic Compute Cloud [Amazon]). We would normally make the storage in EC2, but we need to use Harbor with S3. Because of this, we need to do a lot of things manually. This complicates the processes. The solution is okay, but the issues with storage makes some things a more time-consuming process than they need to be.