Closed szh closed 2 years ago
Note: CI is failing step "OpenShift Oldest 4.x" but this is unrelated and is broken in master as well.
@szh I think the scripts may have assumed that postgres:10
is present in your local registry. Changing the pullPolicy does resolve this , but we don't really want to be automatically searching dockerhub for images if a tag and push step fails during this flow. I think updating 2_prepare_docker_images.sh
with a tag and push step for the postgres image is the way to go, wdyt?
@imheresamir Thanks for the suggestion. I updated and re-committed.
Desired Outcome
Developers should be able to spin up a dev environment on their local machines. Currently this doesn't work because the ImagePullPolicy is set to 'Never' in dev mode, which will cause the script to fail when trying to create a postgres instance because the script doesn't pull the postgres image beforehand.
To reproduce issue: Clone the secrets-provider-for-k8s repo and run
./bin/start --dev --oss
. The script will clone the kubernetes-conjur-deploy repository and attempt to deploy a local cluster, but will timeout when it tries to deploy conjur-postgres. Checking the kubernetes logs, you'll see "Container image "postgres:10" is not present with pull policy of Never". You can confirm the fix by editing thegit clone git@github.com:cyberark/kubernetes-conjur-deploy
line to add-b localbuild
which will use this branch instead of master.Implemented Changes
Connected Issue/Story
N/A
Definition of Done
Changelog
Test coverage
Documentation
README
s) were updated in this PRBehavior
Security