June 2017 Summaries

Git LFS is a Git extension designed to handle large binary files by storing them on an external file server, like JFrog Artifactory, to prevent Git repositories from becoming sluggish due to size inflation. However, storing large files in Artifactory can still lead to storage issues, as each file version is saved in its entirety. To address this, JFrog CLI provides the git-lfs-clean command, which allows for efficient cleanup of outdated or unnecessary binary files from Artifactory Git LFS repositories, thereby freeing up disk space. This command ensures that only the most recent binaries across all branches are retained, with customizable options to specify which files should be preserved or deleted. It is crucial to update all branches before executing the cleanup to avoid accidental deletion of newer files, and users can opt for a dry-run to verify which files will be removed without making any changes.

Gradle's built-in caching feature significantly reduces build times by allowing the sharing of task outputs between machines, enabling subsequent builds to reuse those outputs instead of rebuilding them. This feature supports both local and remote caches, with JFrog Artifactory exemplifying the benefits of a remote cache by reducing build time from 11 seconds to 1 second in a demonstration. The process involves storing the build cache in Artifactory, which can then be accessed by subsequent builds, thereby enhancing efficiency in developer environments. By configuring Gradle to use Artifactory, builds become faster, with Artifactory offering additional benefits such as access control, advanced cleanup, and a distributed cache synchronized across teams. This setup not only accelerates local builds but also speeds up those of remote colleagues, promising significant improvements in build efficiency.

Apache Maven, a prevalent build tool, often utilizes the deploy plugin to upload build artifacts to remote repositories, but it has several limitations such as project aggregation issues, lack of deployment flexibility, and inefficiencies in deployment processes. These drawbacks include inconsistent deployment states when module deployments fail, a rigid approach to artifact deployment, and prolonged cycle times due to unnecessary processing steps. The blog post highlights the importance of context and meta-information in binaries and advocates for decoupling binary management from local build infrastructure to enhance deployment efficiency. It suggests using JFrog Artifactory and its ecosystem, which offer solutions to these challenges, such as atomic transactions, build-tool-agnostic processing, and integrated quality gates. The integration with Jenkins via its Artifactory plugin and the use of Artifactory's REST API or JFrog CLI provides more flexible and maintainable deployment options, allowing users to manage and promote artifacts efficiently, overcoming the limitations of the standard Maven deployment process.