On Saturday, 4 November 2017 17:51:07 UTC-6, Jorge Manrique wrote: > > Hello Andrey, > > POST /service worked, thanks. > > Regarding websockets, I read about the shell/iopub couple here > <https://github.com/sagemath/sagecell/blob/master/doc/messages.md> > I replaced the shell/iopub pair with: > > surl = data.ws_url + "kernel/" + data.id + "/channels"; > try { > wso = new WebSocket(surl); > } catch(e) { > alert(e); > } > > And now the error is: > > WebSocket connection to 'wss:// > sagecell.sagemath.org/kernel/74d6b265-9e9d-4eea-b6a6-ef6d1ace3da1/channels' > failed: Error during WebSocket handshake: Unexpected response code: 403 > > I guess the error is because I do not include the header: > Jupyter-Kernel-ID:" + data.id > when I create the websocket > The issue is that the Javascript Websocket API does not allow to include a > header just like the Python API does > I think SockJS API does not allow it either >
Is it somehow against socket standards that no custom headers can be added? It is used for load balancing and gives much better results than alternatives (which are cookies that tend to create new and new troubles or IP addresses). It is probably possible to add another matching rule like the one for files https://github.com/sagemath/sagecell/blob/master/contrib/vm/container_manager.py#L264 to make all URLs with explicit kernel id match it. > > Any suggestions? > Jorge > > El sábado, 4 de noviembre de 2017, 1:15:10 (UTC-4), Andrey Novoseltsev > escribió: >> >> Hi Jorge, >> >> Thanks a lot for trying and reporting results! >> >> For POST / service: can you try print(factorial(10)) with explicit >> output? I suspect the display hook works differently for it, that's why the >> health check uses code = 'print({} + {})'.format(a, b) >> >> For Websockets: where did you learn shell/iopub additions? They were >> indeed used a couple years ago, but now there is a single link with channel >> name embedded in messages, >> https://github.com/sagemath/sagecell/blob/master/contrib/sagecell-client/sagecell-client.py#L31 >> >> is relevant. >> >> Best, >> Andrey >> > -- You received this message because you are subscribed to the Google Groups "sage-cell" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-cell/549c50c9-310a-4738-9cf6-d34d5607be3e%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
