Description of problem: In the FFWD Upgrade process from Newton->Queens (OSP10->OSP13), there's inconsistency in how the container images file is added to the command line. The converge step seems to need a -e as opposed to using --container-registry-file : openstack overcloud ffwd-upgrade prepare --templates \ --container-registry-file /home/stack/templates/overcloud_images.yaml \ ...<ENV FILES> openstack overcloud ffwd-upgrade converge --templates \ -e /home/stack/templates/overcloud_images.yaml \ ...<ENV FILES> openstack overcloud ceph-upgrade run --templates \ --container-registry-file /home/stack/templates/overcloud_images.yaml \ ...<ENV FILES> It doesn't matter which we choose, '-e' or '--containter-registry-file', let's just make it consistent across the process. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Considering we are out of time for GA, we will propose a more uniform use of the arguments in next release's CLI arguments for FFWD. The most probable case will be the removal of --container-registry-file option in favor of -e, treating it as another environment file.
Created attachment 1446701 [details] how to test https://review.openstack.org/#/q/topic:remove-container-registry-file Adding reviews from https://review.openstack.org/#/q/topic:remove-container-registry-file I sanity checked this on my environment today and ran upgrade prepare/converge without explosions. If you want to test see attachment. The reviews in trackers address this for ceph-upgrade/ffwd-upgrade/upgrade/update but i fear we probably still need to clone this bz for each of those
as per comment #3 and irc yesterday cloned this for the various clis. This one is for ffwd-upgrade so major-upgrade at https://bugzilla.redhat.com/show_bug.cgi?id=1585692 minor update at https://bugzilla.redhat.com/show_bug.cgi?id=1586093 ceph-upgrade at https://bugzilla.redhat.com/show_bug.cgi?id=1586095
Thanks, marios. mcornea alerted me to this change, so I've implemented a fix for it: https://gitlab.cee.redhat.com/rhci-documentation/docs-Red_Hat_Enterprise_Linux_OpenStack_Platform/commit/1a4ec051ee1c53c3291f84f4056efb67b8abed3a Have a look and let me know if this is okay.
thanks Dan! Looks OK, just spotted another nit in the example see https://gitlab.cee.redhat.com/mandreou/docs-Red_Hat_Enterprise_Linux_OpenStack_Platform/commit/e66d99b51b7c800854a01f0d8b492ee9e945e4a9 Also, there will be a warning displayed now that we are no longer using the --container-registry-file and instead pass the container image file with -e. So we should probably include a note about it in the docs? See comment #5 or https://github.com/openstack/python-tripleoclient/blob/74abcc35595d92c39c18d3008ac8df043eff16b6/tripleoclient/utils.py#L984-L990 for the wording of the message that the operator will see. i.e. even though you see the warning, as long as you pass the images with -e it is ok. thanks again.
Thanks, marios. Also spotted another example. Have fixed both and added the note about the message: https://gitlab.cee.redhat.com/rhci-documentation/docs-Red_Hat_Enterprise_Linux_OpenStack_Platform/commit/73ca29073164c1c9e035920cd39f422f17531e03
lgtm but possible s/might/will but up to you ;) thanks very much
Hi, So this can be closed as done as this is exactly the same problem and the same solution than https://bugzilla.redhat.com/show_bug.cgi?id=1586095 and https://bugzilla.redhat.com/show_bug.cgi?id=1585692 and https://bugzilla.redhat.com/show_bug.cgi?id=1586093 I don't know why it has been tagged osp14, but 13 is the place. Moving this to modified, but really this should be closed, the same as the above bz.
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, 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-2019:0045