12-17-2024 02:31 AM
Hi, Please could I ask for some best practices advice regarding Radkit administration.
We need to programatically add users in to the Radkit service, and initially did so using the Radkit Control Api via python scripts. This had some drawbacks due to needing Radkit Control to be updated to keep parity with the service, and adding a lot of libraries to the environment.
We have replicated the needed functionality by directly connecting to the Radkit service (from the same server) using the same http api that the web frontend uses. I didn't see any documentation talking about doing it this way (using chrome debug tools to check the requests and replicating these), so please could we check if this is a supported way of administering Radkit? If so it does look to offer a much more lightweight way to administer the service, however we want to make sure that we are not developing something which may be shut down in the future. If this is an acceptable way to be doing it, are there any plans to add documentation for this method, or is it something for which you would prefer people to be using the control setup? I appreciate that this might be a unique requirement/ setup where the device administering this is the same box from which Radkit service is being run.
Kind regards
Rob
Solved! Go to Solution.
12-20-2024 03:04 AM
Hi Rob,
The ControlAPI is simply a wrapper to the REST API that you are referring to. You can freely use as we do not plan to restrict access.
On the other side, we are also not ready to document it fully yet, so if you can live with the way it is now, we are all good.
The main drawback for you is that REST API changes will be made transparent via the Control API wrapper, but you will have to deal with it directly. Fortunately, we are not planning to modify the core APIs drastically nor often any time soon.
We are however updating the data models (no modification/deletion of parameters, and usually no mandatory additions - only new optional parameters) so be on the lookout. In doubt, you can always use the pydantic models from the Control API to generate sample JSON models and tackle the REST API directly.
Hope this helps!
-- Fred
12-20-2024 03:04 AM
Hi Rob,
The ControlAPI is simply a wrapper to the REST API that you are referring to. You can freely use as we do not plan to restrict access.
On the other side, we are also not ready to document it fully yet, so if you can live with the way it is now, we are all good.
The main drawback for you is that REST API changes will be made transparent via the Control API wrapper, but you will have to deal with it directly. Fortunately, we are not planning to modify the core APIs drastically nor often any time soon.
We are however updating the data models (no modification/deletion of parameters, and usually no mandatory additions - only new optional parameters) so be on the lookout. In doubt, you can always use the pydantic models from the Control API to generate sample JSON models and tackle the REST API directly.
Hope this helps!
-- Fred
01-02-2025 05:28 AM
Hi, sorry for the delay in responding - that's really useful, thank you for your help.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide