Created attachment 1739034 [details] Detailed standalone parameters and deployment action Description of problem: When deploying a standalone environment, system will default to localhost for the undercloud registry instead of the hostname when deploying for the first time. When redeploying the UC ctlplane hostname will have been added to /etc/hosts so it will use the correct hostname when attempting to access the undercloud container regsitry TASK [include_tasks] ********************************************************************************************************************************************************************** Monday 14 December 2020 12:08:45 -0500 (0:00:00.094) 0:03:57.194 ******* included: /home/stack/standalone-ansible-gqj_bg07/Standalone/deploy_steps_tasks_step1.yaml for intel-purley-lr-04 TASK [tripleo-container-tag : Fail if container image is undefined] *********************************************************************************************************************** Monday 14 December 2020 12:08:45 -0500 (0:00:00.159) 0:03:57.353 ******* skipping: [intel-purley-lr-04] TASK [tripleo-container-tag : Pull localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1 image] ************************************************************************** Monday 14 December 2020 12:08:45 -0500 (0:00:00.096) 0:03:57.450 ******* fatal: [intel-purley-lr-04]: FAILED! => {"changed": true, "cmd": "podman pull localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1", "delta": "0:00:00.064277", "end": "2020-12-14 17:08:45.857774", "msg": "non-zero return code", "rc": 125, "start": "2020-12-14 17:08:45.793497", "stderr": "Trying to pull localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1...\n Get https://localhost:8787/v2/: http: server gave HTTP response to HTTPS client\nError: error pulling image \"localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1\": unable to pull localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1: unable to pull image: Error initializing source docker://localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1: error pinging docker registry localhost:8787: Get https://localhost:8787/v2/: http: server gave HTTP response to HTTPS client", "stderr_lines": ["Trying to pull localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1...", " Get https://localhost:8787/v2/: http: server gave HTTP response to HTTPS client", "Error: error pulling image \"localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1\": unable to pull localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1: unable to pull image: Error initializing source docker://localhost:8787/rh-osbs/rhosp16-openstack-cinder-volume:16.1_20201202.1: error pinging docker registry localhost:8787: Get https://localhost:8787/v2/: http: server gave HTTP response to HTTPS client"], "stdout": "", "stdout_lines": []} NO MORE HOSTS LEFT ************************************************************************************************************************************************************************ PLAY RECAP ******************************************************************************************************************************************************************************** intel-purley-lr-04 : ok=170 changed=106 unreachable=0 failed=1 skipped=41 rescued=0 ignored=0 undercloud : ok=23 changed=15 unreachable=0 failed=0 skipped=1 rescued=0 ignored=0 Version-Release number of selected component (if applicable): 16.1 How reproducible: 100% on fresh deployment attempt, second attempt will not have the issue Steps to Reproduce: Full details in attachment (standalone_deployment_failure) 1. Deploy standalone fresh standalone deployment. 2. 3. Actual results: Deployment attempts to use localhost when accessing container registry Expected results: Standalone (All-in-one) deployment uses the hostname when attempting to access the undercloud container registry Additional info:
Do not use push_destination: true in the container-image-prepare.yaml when deploying the all-in-one.
Hello James, Would it be possible to provide the content of the /etc/hosts? Apparently there is a detection being done that wants to match the hostname on the ctlplane network, maybe it's just missing this bit (even for standalone).. Cheers, C.
Hi Cédric, Correct the ctlplane mapping is not present in the hosts file whenever I attempt to deploy for the first time on a newly provisioned beaker host. The subsequent deployment attempt will work since the /etc/hosts file is updated with the ctlplane network info. Info listed below. Regards, -James # Before deploying [root@intel-purley-lr-04 ~]# cat /etc/hosts 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6 # During first deploy attempt ...... jinja2 rendering roles Standalone rendering j2 template to file: /home/stack/tripleo-heat-installer-templates/./puppet/standalone-role.yaml No entry for intel-purley-lr-04.ctlplane in /etc/hosts. Falling back to use the default (localhost) undercloud registry. No entry for intel-purley-lr-04.ctlplane in /etc/hosts. Falling back to use the default (localhost) undercloud registry. ** Performing Heat stack create.. ** ..... Exception: Deployment failed Traceback (most recent call last): File "/usr/lib/python3.6/site-packages/tripleoclient/v1/tripleo_deploy.py", line 1328, in _standalone_deploy raise exceptions.DeploymentError('Deployment failed') tripleoclient.exceptions.DeploymentError: Deployment failed None Not cleaning working directory /home/stack/tripleo-heat-installer-templates Not cleaning ansible directory /home/stack/standalone-ansible-gqj_bg07 Install artifact is located at /home/stack/standalone-install-20201214170846.tar.bzip2 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Deployment Failed! ERROR: Heat log files: /var/log/heat-launcher/undercloud_deploy-bza_001b !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Deployment failed. # /etc/hosts has been updated with ctlplane mappings and second attempt to deploy will now work [stack@intel-purley-lr-04 ~]$ cat /etc/hosts # BEGIN ANSIBLE MANAGED BLOCK 192.168.24.101 intel-purley-lr-04.localdomain intel-purley-lr-04 192.168.24.101 intel-purley-lr-04.ctlplane.localdomain intel-purley-lr-04.ctlplane # END ANSIBLE MANAGED BLOCK 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6 10.46.9.32 rhos-qe-mirror-tlv.usersys.redhat.com download.lab.bos.redhat.com download.eng.bos.redhat.com download-node-02.eng.bos.redhat.com
Moving to OSP17 for now as we can probably address it for there. Given the standalone configuration is mostly for development purposes, we probably won't backport to 16.1.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Release of components for Red Hat OpenStack Platform 17.0 (Wallaby)), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2022:6543