Galaxy deployment via Tools Hub on the EOSC EU Node

[sebastian-luna-valero]

Hello,

@kotsu21, @bgruening, and I are testing the TOSCA templates for Galaxy in the EOSC EU node. We would like to showcase this in the upcoming EOSC Winter School as part of the Opportunity Area 4: https://eosc.eu/eosc-focus-project/winter-school-2025/

A deployment of the below template does not seem to work:
https://github.com/grycap/tosca/blob/eosc_lot1/templates/galaxy.yaml

Via the Tools Hub on the EOSC EU Node:
https://open-science-cloud.ec.europa.eu/dashboard/tools-hub/tool-view?eo=MjEuMTExNjcvQnVZYk53

It creates this endpoint that seems inaccessible: https://62.3.174.204/galaxy

Has this TOSCA template been tested successfully in the past? or does it need an update?

Best regards,
Sebastian

[micafer]

Hi @sebastian-luna-valero,

The problem with this recipe is that it takes some time to be up, after the infrastructure has been configured.
Could you try again and wait some minutes to ensure that all the services have completed the initialization steps?

[kotsu21]

As you say, the deployment took some time but generated the output endpoint https://62.3.174.204/galaxy (it may have been expired already as it was created for testing with the minimum credits).
By the way, should my OS be Linux to access this endpoint?

[micafer]

As you say, the deployment took some time but generated the output endpoint https://62.3.174.204/galaxy (it may have been expired already as it was created for testing with the minimum credits). By the way, should my OS be Linux to access this endpoint?

Yo only need a browser, it can be used by any OS,
In this case it seems that there is a misconfiguration.
I will test it again.

[bgruening]

I think we tried 5min later or so and we still just saw the nginx error page.
Which version of the container is used here? The new version 24.1 should have a nicer nginx error page.

Thanks @micafer for looking at this.

[micafer]

I updated the version of the helm chart and now it deploys last Galaxy version.
But there is no nice error page ...
The problem is that the amount of resources set in the Galaxy EU Node template are to low.
You could try creating a copy of the Galaxy template in your tools, and increase the number or resources.
You only need to update the input values.
The recommended option is to set 3 WNs with 2 CPUs and 8 GBs or RAM.
Please test it and let me know.

[kotsu21]

Thanks for the update. I changed the input values (see below) and it took 20 minutes to complete deploying. The endpoint is still not accessible via https://62.3.174.144/galaxy. What interface are we expecting to see?
wn_num:
type: integer
description: Number of WNs in the cluster
default: 3
required: yes
wn_cpus:
type: integer
description: Number of CPUs for the WNs
default: 2
required: yes
wn_mem:
type: scalar-unit.size
description: Amount of Memory for the WNs
default: 8 GB
required: yes
When I clicked the deployed VM to view externally from the EU Node platform, api.cloud.psnc.pl was accessible.