Port not handled properly in containerized image references

[vladimir-mzv]

Hi,
I'm trying to upgrade from Murphy Appliance to Oparin k8s. After installing Oparin on a fresh k8s 1.26, and restoring old PG database, operator pod starts successfully, but all the autogenerated deployments cannot start due to incorrectly set image:

1-generic-7f456fb8d8-9b9kj                                    0/1     InvalidImageName    0          7s
1-generic-7f456fb8d8-prglr                                    0/1     InvalidImageName    0          7s
1-priority-55fc744578-92w2h                                   0/1     InvalidImageName    0          7s
1-priority-55fc744578-lj6n5                                   0/1     InvalidImageName    0          6s
1-schedule-758987d86-5psbs                                    0/1     InvalidImageName    0          6s
2-ansible-tower-automation-event-catcher-4-5bb4d59df6-6c87g   0/1     InvalidImageName    0          29s
2-ansible-tower-automation-refresh-4-5778799785-t9g75         0/1     InvalidImageName    0          24s
2-ems-metrics-processor-5877d4f85f-2f4pj                      0/1     InvalidImageName    0          23s
2-ems-metrics-processor-5877d4f85f-rtdmg                      0/1     InvalidImageName    0          23s
2-event-handler-78c6785f48-skppj                              0/1     InvalidImageName    0          23s
2-generic-88d758945-52ndf                                     0/1     InvalidImageName    0          22s
2-generic-88d758945-lrrx8                                     0/1     InvalidImageName    0          23s
2-priority-64dfb5cff7-6qvvf                                   0/1     InvalidImageName    0          22s
2-priority-64dfb5cff7-g7zd7                                   0/1     InvalidImageName    0          22s
2-remote-console-6c9bc5d6d9-hf49d                             0/1     Running             0          21s
2-reporting-85cd8b8bcc-srr55                                  0/1     InvalidImageName    0          22s
2-reporting-85cd8b8bcc-x2x9j                                  0/1     InvalidImageName    0          21s
2-schedule-cddf844f-9sf7t                                     0/1     InvalidImageName    0          21s
2-ui-7ccbb486bf-6q8tp                                         0/1     ContainerCreating   0          20s
2-ui-7ccbb486bf-jzcnz                                         0/1     Running             0          20s
2-web-service-b58f686ff-c8pl9                                 0/1     Running             0          20s
httpd-84b9958c84-nx2sz                                        1/1     Running             0          51m
manageiq-operator-7b87f87db6-d4cpf                            1/1     Running             0          51m
memcached-787c4698bf-28b2s                                    1/1     Running             0          51m
orchestrator-5cc99d88b9-j7v5h                                 1/1     Running             0          2m5s
postgresql-666cdcbb78-sfpkw                                   1/1     Running             0          51m

the image for them looks like this:
artifactory.my.local/manageiq-base-worker:17011/manageiq/manageiq-orchestrator

Probably it is related to env variables set in orchestrator deployment:

CONTAINER_IMAGE_NAMESPACE:  artifactory.my.local
CONTAINER_IMAGE_TAG:        17011/manageiq/manageiq-orchestrator

but orchestrator itself takes correct image address given in CR:
orchestratorImage: artifactory.my.local:17011/manageiq/manageiq-orchestrator:oparin-1.1

How to apply correct image for autogenerated pods?

[Fryguy]

Did you set the GUID in the CR to the same GUID from the appliance? Can you show your CR?

[Fryguy]

artifactory.my.local is wrong (unless that's your artifactory instance?)

@bdunne @nasark can you look here?

[vladimir-mzv]

Sure, it's an internal repo.

spec:
  appName: manageiq
  applicationDomain: manageiq.my.local
  backupLabelName: manageiq.org/backup
  databaseRegion: "0"
  databaseSecret: postgresql-secrets
  databaseVolumeCapacity: 15Gi
  deployMessagingService: false
  enableApplicationLocalLogin: true
  enableSSO: false
  enforceWorkerResourceConstraints: false
  httpdAuthenticationType: internal
  httpdImage: artifactory.my.local:17011/manageiq/httpd:latest_01242024
  kafkaImage: artifactory.my.local:17011/manageiq/kafka:latest_01242024
  kafkaVolumeCapacity: 2Gi
  memcachedImage: artifactory.my.local:17011/manageiq/memcached:1.6
  memcachedMaxConnection: "1024"
  memcachedMaxMemory: "64"
  memcachedSlabPageSize: 1m
  migrationsRan:
  - "20210503163000"
  - "20210504113000"
  orchestratorImage: artifactory.my.local:17011/manageiq/manageiq-orchestrator:oparin-1.1
  orchestratorInitialDelay: "480"
  postgresqlImage: artifactory.my.local:17011/manageiq/postgresql:13
  postgresqlMaxConnections: "1000"
  postgresqlSharedBuffers: 1GB
  serverGuid: bd87dfc5-8066-4f07-b0cb-9105965c8b0a
  uiWorkerImage: artifactory.my.local:17011/manageiq/manageiq-ui-worker:oparin-1.1
  webserverWorkerImage: artifactory.my.local:17011/manageiq/manageiq-base-worker:oparin-1.1
  zookeeperImage: artifactory.my.local:17011/manageiq/zookeeper:latest_01242024
  zookeeperVolumeCapacity: 1Gi

[Fryguy]

Ok, so we are clearly not handling the port in the image name, and instead see the : as the image tag separator. @bdunne This looks like a clear bug to me. Please take a look.

[miq-bot]

This issue has been automatically marked as stale because it has not been updated for at least 3 months.

If you can still reproduce this issue on the current release or on master, please reply with all of the information you have about it in order to keep the issue open.