No css on OMERO.web about ansible-example-omero-onenode HOT 9 CLOSED

One idea:

Many playbooks in https://github.com/ome/prod-playbooks have nginx capped at 1.18.0, such as https://github.com/ome/prod-playbooks/blob/3b3d2e0f2f1f048b2219c9094d7bef8cb43ae6fc/web-proxy/playbook.yml#L32.
But in my workflow, I am having nginx 1.22.x.
I suppose that the capping is done in a variable which is used by the role ome.nginx ?

from ansible-example-omero-onenode.

In the OMEROweb.log I can see

2022-09-02 14:47:27,692 WARNI [                          django.request] (proc.06290) log_response():230 HTTP 404 <WSGIRequest: GET '/static/webclient/image/custom_login_logo.png'>

I think that this is the reason that these css and other static files are not served and I cannot see them in the webclient login page. I tried to

1. compare the permissions on the folders such as /static/webclient/image/ - the /static folder actually did not exist, but this is in line with the functioning outreach server
2. I created the /static/webclient/image/ and other folders by hand and copy the files from /opt/omero/web/OMERO.web/var/static/ to the newly manually created /static and restart nginx and OMERO.web, and it did not help (still no css on login page)
3. I compared the perms on the /opt/omero/web/OMERO.web/var/static/ folders in comparison with outreach -> no differences detected

from ansible-example-omero-onenode.

Is this related to @jburel's recent whitenoise PR?

from ansible-example-omero-onenode.

this is more likely to be related to a omero.web configuration. something like omero config set omero.web.debug True might help cc @pwalczysko

from ansible-example-omero-onenode.

Thanks a lot @jburel , this fixed it.
I tested it yesterday manually fixing the config and the css appeared as expected.
I am now working on the full workflow using ansible, and will open a PR fixing the playbook.

from ansible-example-omero-onenode.

Closed #16 because, although it fixed the problem, it was adding unnecessary config steps.

The workflow here works out of the box, the only gotcha was to add a port-forwarding for

-p 8080:80

into the docker run... command (edited in the header of this issue) and then go to web on http://localhost:8080 and all works fine.

from ansible-example-omero-onenode.

Tried following steps above and got this error (on my M1 Mac):

$ docker run --privileged --name httpd -p 22 -p 4063:4063 -p 4064:4064 -p 4080:4080 -p 8080:80 -d httpd
WARNING: The requested image's platform (linux/amd64) does not match the detected host platform (linux/arm64/v8) and no specific platform was requested
c64daacc3048e4985829d07f3ed8d7318bf20662f3591ee82224ecf72e6e9eed

from ansible-example-omero-onenode.

@will-moore Strangely, I was able to go over the docker run... step on my M1 machine successfully (I was successful already previously on the MacBookPro).
Do you want to try some solutions from https://stackoverflow.com/questions/66662820/m1-docker-preview-and-keycloak-images-platform-linux-amd64-does-not-match-th ?
The docker build -t <image-name> --platform linux/x86_64 which you would add to the previous build command seems low cost to try and fairly logical to me ?

from ansible-example-omero-onenode.

Tried:

$ docker build -f Dockerfile.j2 --rm --no-cache -t httpd --platform linux/x86_64 .

$ docker run --privileged --name httpd -p 22 -p 4063:4063 -p 4064:4064 -p 4080:4080 -p 8080:80 -d httpd
docker: Error response from daemon: Conflict. The container name "/httpd" is already in use by container...

$ docker run --privileged --name httpd2 -p 22 -p 4063:4063 -p 4064:4064 -p 4080:4080 -p 8080:80 -d httpd
WARNING: The requested image's platform (linux/amd64) does not match the detected host platform (linux/arm64/v8) and no specific platform was requested
d1c2aa3a1eaf81d5e2703fbf1f0607c5da44efbd954c7647a17792469719d1ad

$ docker exec -it httpd2 bash
[root@d1c2aa3a1eaf /]#

# followed all the steps in description above...

...until...

[root@d1c2aa3a1eaf /]# ansible-playbook --become --connection=local -i hosts.yml playbook.yml
...
TASK [ome.postgresql : postgres | initialise PostgreSQL cluster (skip if data directory already exists)] ******************************************
fatal: [localhost]: FAILED! => {"changed": true, "cmd": ["/usr/pgsql-13/bin/postgresql-13-setup", "initdb"], "delta": "0:00:00.514084", "end": "2022-11-04 10:06:09.047965", "msg": "non-zero return code", "rc": 1, "start": "2022-11-04 10:06:08.533881", "stderr": "Failed to get D-Bus connection: No such file or directory", "stderr_lines": ["Failed to get D-Bus connection: No such file or directory"], "stdout": "failed to find PGDATA setting in postgresql-13.service", "stdout_lines": ["failed to find PGDATA setting in postgresql-13.service"]}

PLAY RECAP ****************************************************************************************************************************************
localhost                  : ok=8    changed=5    unreachable=0    failed=1    skipped=11   rescued=0    ignored=0   

from ansible-example-omero-onenode.