Don't know. Slurm-web don't work at all.

[BlackS52]

Wanted to install it to SUSE. Had many problems with different /etc/passwd, users www_data, _apt and etc. Umounted
Found that the docker file tried to install SLURM verstion 15 instead current v22. Fixed it.
Found munge problem when a spawned a new munge process each 1 sec that lead to crash your system. Fixed it in /etc/services/munge/run

But slurm-web still don't show anything in web browser. Only one button and nothing more.

Why ? Don't know. Looks like a project is died. Hope that I wrong.
The great idea but don't worked realisation.
Sad.

[wangzhen2271]

I have the same problem.

[carlos-encs]

Same issue here.
Need some help.

[hellozzcy]

Same issue here.
Need some help.

[rezib]

The problem is probably due to a background HTTP request with an unexpected result. Can you please check the network requests in your browser development console and find out what is going wrong?

[rezib]

Some documentation for this configuration file is available here:
http://rackslab.github.io/slurm-web/installation.html#clusters-clusters-config-json

There are two HTTP WSGI applications involved in Slurm-web : confdashboard and restapi. The restapi must be deployed on all clusters, the confdashboard must be deployed on a central server with the HTML page and the JS code. In a configuration with one local Slurm cluster, all these components can be deployed on one server.

The first application confdashboard basically provides cluster.conf.json which contains the list of Slurm clusters, with the URL to access their corresponding restapi components. The URL value depends on the way the restapi application is launched (eg. with apache/nginx or as standalone Flask application), its server and listening port.

[hellozzcy]

I am deploying slurm-web using Docker. How can I view the restapi configuration inside Docker? I didn't see any introduction on the official website. Also, the confdashboard you mentioned, how is it configured to read the local slurm cluster? I also didn't see any introduction in the official documentation. I would appreciate your answers, thank you very much.

[rezib]

Yes, the Docker container was an external contribution and it misses information in the existing documentation. The configuration file in the container is binded to the conf\ subdirectory in the current working:

https://github.com/rackslab/slurm-web/blob/90463564375148b5c67e485006c6e2a4da5eb05e/docker/run/run.sh#L4

The default empty array in clusters.conf.json should be fine in this configuration, as it considers the restapi component to be available under the slurm-restapi path on the same HTTP server. The problem you are facing is probably not caused by a misconfigured clusters.conf.json but more due to the restapi encountering an undefined issue leading to HTTP/500 response codes. You should probably dig into the logs of the restapi component to discover what is going wrong there.

[hellozzcy]

Is there any dependency between the authorized_origins configuration items in the cluster.conf.json and restapi.conf files, and I checked the logs, the command used is docker logs -f XXXX, I did not find any errors, I want to find inside docker How can I check the running status of these two applications, restapi and confdashboard, I want to check their configuration items, I really need your answer Thanks

[rezib]

I guess you can find some interesting stuff in /var/log/apache inside the container, by opening an interactive shell with docker.

[rezib]

This question was about deprecated Slurm-web 2, then I close the discussion.