ClearML FAQ | Hi All! I Have A Self-Hosted Clearml Instance Running On My Server. I See This In The Docs: """ - Or Manually Edit `~/Clearml.Conf` File, Making Sure The Server Settings (`Api_Server`, `Web_Server`, `File_Server`) Are Configured Correctly, For Example:

Answered

Hi All! I Have A Self-Hosted Clearml Instance Running On My Server. I See This In The Docs: """ - Or Manually Edit `~/Clearml.Conf` File, Making Sure The Server Settings (`Api_Server`, `Web_Server`, `File_Server`) Are Configured Correctly, For Example:

Hi all!
I have a self-hosted ClearML instance running on my server. I see this in the docs:
"""

Or manually edit ~/clearml.conf file, making sure the server settings (api_server, web_server, file_server) are configured correctly, for example:

        api {
            # API server on port 8008
            api_server: "

"

            # web_server on port 8080
            web_server: "

"

            # file server on port 8081
            files_server: "

"
        }

"""

To expose it to the internet and communicate with the server via the ClearML SDK, these endpoints don't work. I'm routing requests to / to localhost:8080 , /api to localhost:/008 , and /files to localhost:8081 on my server. Hence, a working client side clearml.conf looks like this:

        api {
            api_server: "

"
            web_server: "

"
            files_server: "

"
        }

The generated ClearML Credentials from the webUI have to be modified from containing the original ports to the aforementioned example. I am wondering if I can configure some environment variable or modify some build-time settings with my own base url?

  				
Posted 
	9 months ago

					More
				  		
  Report
		
					HighCoyote66
				
					0
					 × 1

Votes Newest

Answers 2

Hi @<1600661428556009472:profile|HighCoyote66> , the UI should be able to generate these URLs from the location it sees in the browser, what address are you using to browse to the UI?

  				
Posted 
	9 months ago

					More
				  		
  Report
		
					SuccessfulKoala55
				
					0
					 × 1

Hey @<1523701087100473344:profile|SuccessfulKoala55> thanks for the reply and apologies for the delayed response. I added an nginx webserver on top of the default open-source clearml to route requests from external clients to the host machine's ports 8080, 8008, and 8081. To browse to the UI, I use my personal domain (which resolves to the machines static IP), e.g. None I've redacted it here for privacy if you don't mind.

The only port exposed to the internet is port 80 which my nginx reverse proxy is listening to. From there, I just route requests made to /api to 8008, /files to 8081, and requests to the default webserver to 8080. I hope this makes sense I'm not the best at explaining 😓 .

With this config, the SDK works fine. I'm able to upload stuff to the fileserver, access the API, all the basic features. I was more wondering if I'm missing something in the docs where I can tell the application what is being exposed, so that the ClearML credentials generate correctly. This might also cover some niche use cases where these key: value pairs need to be accessed internally by the app itself.

  				
Posted 
	9 months ago

					More
				  		
  Report
		
					HighCoyote66
				
					0
					 × 1

Write your answer

656 Views

2 Answers

9 months ago