josephevans commented on a change in pull request #19654:
URL: https://github.com/apache/incubator-mxnet/pull/19654#discussion_r541172878
##########
File path: ci/build.py
##########
@@ -117,6 +123,9 @@ def run_cmd():
image_id = _get_local_image_id(docker_tag=tag)
if not image_id:
raise FileNotFoundError('Unable to find docker image id matching with
{}'.format(tag))
+ # now that we've built the container, push it to our docker cache if
DOCKER_ECR_CACHE is defined
+ if 'DOCKER_ECR_REGISTRY' in os.environ:
+ push_docker_cache(registry, tag, image_id)
Review comment:
Yes, it allows us to cache much more aggressively and eliminate the need
for a job to build and push these images up. The ECR registry is not public
readable, so it's not like we could potentially publish something bad. IMO,
this is no different than using ccache for compiled files.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
