Created attachment 1157644 [details] evm.log attached. Description of problem: Run smart state analysis on any container image Version-Release number of selected component (if applicable): 5.6. How reproducible: Always Steps to Reproduce: 1.Select container - container images 1.1. select a container image 2.Select Configuration - Smart State Analysis 3.Go to Tasks - All VM and container analysis tasks Actual results: Error is displayed: Cannot analyze non docker images Expected results: Task should be completed succesfully and package information should be provided for the selected container Additional info: EVM.log attached
Connecting to the the appliance I can see that one of the jobs for one of the containers failed with "Cannot analyze non docker images". Inspecting that image, I can see it cannot grab the image, since the downstream image-inspector is not yet available[1] Please follow the following step until then[2] The other scans failed due to openshift not being reachable [1] :containerStatuses=>[{:name=>"image-inspector", :state=>{:waiting=>{:reason=>"ImagePullBackOff", :message=>"Back-off pulling image \"registry.access.redhat.com/openshift3/image-inspector:2.0.0\""}}, :lastState=>{}, :ready=>false, :restartCount=>0, :image=>"registry.access.redhat.com/openshift3/image-inspector:2.0.0", :imageID=>""} [2] Edit /var/www/miq/vmdb/config/initializers/override_containers_scanning_image.rb Change 'registry.access.redhat.com/openshift3/image-inspector' to 'docker.io/openshift/image-inspector:v2.0.z' restart the service
(In reply to Mooli Tayer from comment #2) > Connecting to the the appliance I can see that one of the jobs for one of > the containers failed with "Cannot analyze non docker images". Mooli, isn't the error message misleading? If it's impossible to pull the image, can't we make the error more explanatory?
Sure we can improve. I wonder if this is an acceptable way: during ssa if docker_id is empty && image_ref is empty simply give a different message, e.g "Image not available"? Another issue I noticed is that for both these errors[1] we should also log them, like we log other errors. I will submit a patch. [1] https://github.com/ManageIQ/manageiq/blob/master/app/models/manageiq/providers/kubernetes/container_manager/scanning/job.rb#L39