tl;dr: The artifacts do in fact upload but I am having trouble finding the bottleneck. Publishing the 2.3GB artifact from a Jenkins docker container to a co-located Nexus docker container is taking an average of 20 minutes.
Long version:
I have a single box running two Docker containers, one for Jenkins and one for Nexus 3 OSS. Both Docker containers have their own co-located volumes attached for persistent storage. The Jenkins instance (that is running in Docker) has the Nexus plugin installed. Jenkins builds are a multi-branch pipeline orchestrated using a jenkinsfile. Nexus is set up with a maven repo to receive the artifacts from Jenkins when it reaches the publish step and invokes nexusPublisher. It takes an average of 20 minutes to push the 2.3GB artifact out of the Jenkins container/volume and into the co-located Nexus container/volume.
I am struggling to understand how two container running on the same box are slow to essentially move the artifact on the same underlying hard drive. Granted, Jenkins is pushing out of its container and into the Nexus container via HTTP, but with them co-located I would have expected faster thruput.
Any insight into how to speed this up or things to check would be much appreciated.