Description
An Artifactory server I manage via this Ansible collection just recently experienced issues with artifacts larger than 1GB being cutoff at the 1GB mark during their downloads.
After reviewing all the settings in Artifactory itself, and it's configuration, along with the AWS ELB, we made a change in the nginx configuration to the proxy_max_temp_file_size
setting. This defaults to 1024m
and was the cause of the artifact downloads not finishing all at once. Certain clients were able to pick up where the download had left off, but this caused several CI/CD jobs to fail.
Considering how the client_max_body_size 0
in the artifactory.conf
nginx template, I don't think there'd be a problem setting proxy_max_temp_file_size 0
in there as well, and would allow the larger artifact downloads to not be interrupted.
We are utilizing the S3 artifact storage, if that makes any difference.