Does it progress any further if the repo username and repo are correct? You said it was failing when attempting to clone the repo (which makes sense since the user was wrong).
Edit: by the way, I looked at the recent role import failures and I see a bunch of variations you’ve tried, but I don’t see any using eclipse-slm as the Github user, so it might be worth trying again.
by the way, I looked at the recent role import failures and I see a bunch of variations you’ve tried, but I don’t see any using eclipse-slm as the Github user, so it might be worth trying again.
With trying:
ansible-galaxy role import --token $AG_TOKEN eclipse-slm slm-ansible-role-docker
and this failing with Status Code 500, I tried using “eclipse-slm” as github user right? Or do i misunderstand something at this point? The tries with “eclipse-slm” do not show up in the galaxy task list because they do not even initiate the import process and consequently they do not show up at all.
Does it progress any further if the repo username and repo are correct? You said it was failing when attempting to clone the repo (which makes sense since the user was wrong).
If i use “ipa-big” (my github username) or “eclipse_slm” (the galaxy namespace - as “eclipse-slm” the real github org name is not possible) the import process starts but fails when cloning. When using “eclipse-slm” the import process doesnt even start in first place.
Since you have the role_name and namespace defined in the meta/main.yml , I’d expect the real Github user/repo to be passed to ansible-galaxy .
As you mentioned the meta/main.yml i now remove namespace and role_name from meta and retried:
ansible-galaxy role import --token $AG_TOKEN ipa-big slm-ansible-role-docker
# and:
ansible-galaxy role import --token $AG_TOKEN eclipse_slm slm-ansible-role-docker
tippe oder füge den Code hier ein
… but still get the same result:
Starting import: task_id=2148100314115555454674309565326574450, pulp_id=019db585-bb97-74d1-81cb-886b81284372
==== PARAMETERS ====
importer username: ipa-big
matched user: ipa-big id:3120
github_user: eclipse_slm
github_repo: slm-ansible-role-docker
github_reference: None
alternate_clone_url: None
alternate_namespace_name: None
alternate_role_name: None
==== CHECK FOR MATCHING ROLE(S) ====
user:eclipse_slm repo:slm-ansible-role-docker did not match any existing roles
===== CLONING REPO =====
cloning https://github.com/eclipse_slm/slm-ansible-role-docker ...
cloning failed: Cloning into '/tmp/tmpkcvbq6ib/slm-ansible-role-docker'...
fatal: could not read Username for 'https://github.com': terminal prompts disabled
I should have been clearer, but if you need to upload a role to a different namespace, you MUST provide the namespace in the meta/main.yml. You can provide a different role name on the CLI, but there’s no corresponding toggle for the namespace.
With trying:
ansible-galaxy role import --token $AG_TOKEN eclipse-slm slm-ansible-role-docker
and this failing with Status Code 500, I tried using “eclipse-slm” as github user right? Or do i misunderstand something at this point? The tries with “eclipse-slm” do not show up in the galaxy task list because they do not even initiate the import process and consequently they do not show up at all.
Yes, that’s the command I was asking you to reattempt. The positional arguments are used to compose the Github repo to clone. I have no idea what you mean that it doesn’t show up at all, since you’ve only shared an error with no context. Can you run it with more verbosity -vvvv and share the full output instead of the snippet?
Hey! I’ve seen similar issues when the GitHub organization’s third-party application restrictions are turned on. Even if you’re an owner, you sometimes have to explicitly ‘Grant’ access to the Ansible Galaxy app within the organization settings on GitHub for it to actually see the repos.
Another thing to double-check is that your meta/main.yml has the correct namespace and role name matching the new org exactly. Sometimes Galaxy gets hung up on the old cache if there’s even a tiny mismatch in the naming convention.
Can you run it with more verbosity -vvvv and share the full output instead of the snippet?
Running this (adding -vvvv):
ansible-galaxy role import -vvvv --token $AG_TOKEN eclipse-slm slm-ansible-role-docker
return this:
ansible-galaxy [core 2.20.4]
config file = None
configured module search path = ['/root/.ansible/plugins/modules', '/usr/share/ansible/plugins/modules']
ansible python module location = /usr/local/lib/python3.12/site-packages/ansible
ansible collection location = /root/.ansible/collections:/usr/share/ansible/collections
executable location = /usr/local/bin/ansible-galaxy
python version = 3.12.13 (main, Apr 7 2026, 04:52:21) [GCC 14.2.0] (/usr/local/bin/python3.12)
jinja version = 3.1.6
pyyaml version = 6.0.3 (with libyaml v0.2.5)
No config file found; using defaults
Initial connection to galaxy_server: https://galaxy.ansible.com
Opened /root/.ansible/galaxy_token
Calling Galaxy at https://galaxy.ansible.com/api/
Found API version 'v3, pulp-v3, v1' with Galaxy server default (https://galaxy.ansible.com/api/)
Calling Galaxy at https://galaxy.ansible.com/api/v1/imports/
[ERROR]: None (HTTP Code: 500, Message: Internal Server Error)
… and regarding this:
I have no idea what you mean that it doesn’t show up at all, since you’ve only shared an error with no context.
… i mean that when i run the ansible-galaxy import command mentioned before, which gives me status code 500, this import try does not show up in the ansible ui at Ansible Galaxy
Even if you’re an owner, you sometimes have to explicitly ‘Grant’ access to the Ansible Galaxy app within the organization settings on GitHub for it to actually see the repos.
According to this: Create a namespace for organization · Issue #1315 · ansible/galaxy · GitHub I now found out that Ansible Galaxy has no full access to the github organisation. As the organization “eclipse-slm” is an eclipse-owned org and i’m just a contributer to this org i have to send a request for full access to the org - what i did now. This will take a while i guess.
Another thing to double-check is that your meta/main.yml has the correct namespace and role name matching the new org exactly.
