Help required to resolve the Websocket connection failed error with a docker image with separate Trame Server + Vanilla JS frontend client

[topcat]

Hello,

I would like to use Trame as the server with a separate vanilla JS frontend client. I would like both the server and client to be hosted separately as well. The client would need to create a Websocket connection to the server.

I came across mention from @jourdain in another thread of the example https://github.com/Kitware/vtk-web-solutions where you have a react and vue client with a backend trame server.

Taking the above as a basis, I have converted the code from React / Vue as a frontend to vanilla JS. I have used the trame-cookiecutter to create the trame server and the docker images for these.

I have deployed the docker server onto AWS Lightsail and I have hosted the client on S3 to test all of this out.

My JS client attempts to create a Websocket link to the server and when it does that it fails with the following in my console log in the browser :

WebSocket connection to 'ws://[path to my container public endpoint]/ws' failed:

I am very new to Trame and probably need to do something to enable websocket forwarding in the docker image and / or open up some ports - currently my container is listening in only on port 80.

Would appreciate help on how I could go about debugging this.

Thank you for your time.

Thank you.

Regards,
Jeetu

[jourdain]

That behavior is expected because you never started a session. Also that websocket endpoint is wrong for a mulit-user/docker setup.
But guiding you through a discussion will be tricky, there will be too much to cover. You can obviously go through the documentation of the launcher to understand the principle but it might be easier to just grab a small support with Kitware and then we can walk you through how you can leverage trame in your specific setup.

[topcat]

Hi Sébastien, Thank you for your response. Currently I am evaluating Trame for a particular biomedical research project for a centre at the Indian Institute of Technology, Bombay. The project is not funded at the moment and we are attempting to create a MVP that we can then use to get funds / grants. Therefore while at the moment we will not be able to engage with Kitware commercially I do see such a synergy possible in the future since the overall project vision is very substantial. I understand your difficulty in covering all of this well enough for a trame novice to grasp here, however if you could give me any pointers, examples and resources rhen I would greatly appreciate it. It would be great if we could use Trame for this Work since I really like what I have seen of it thus far. Thank you again. Regards, Jeetu

…

On Wed, 15 May, 2024, 00:56 Sebastien Jourdain (Kitware), < ***@***.***> wrote: That behavior is expected because you never started a session. Also that websocket endpoint is wrong for a mulit-user/docker setup. But guiding you through a discussion will be tricky, there will be too much to cover. You can obviously go through the documentation of the launcher to understand the principle but it might be easier to just grab a small support with Kitware and then we can walk you through how you can leverage trame in your specific setup. — Reply to this email directly, view it on GitHub <#523 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNNFZFDD2WUIE7NMYATZCJQNXAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIMZXG43TS> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jourdain]

As a first step, I would just drop your react app into /setup/www so it can be bundled with the regular trame docker step.
Then make sure your wslink config for setting up the connection only has {application: 'trame'}.
That should be enough to fix those session/websocket issue.

In another discussion I've listed all the doc on it but I can't seem to find it again.

[jourdain]

But you can look at that one along with a couple of others on that same website.

[topcat]

Hi Sebastien, Thank you for the pointers and the documents. I shall attempt to follow suit with what you have suggested. I would appreciate a few clarifications below : <quote> As a first step, I would just drop your react app into /setup/www so it can be bundled with the regular trame docker step. Then make sure your wslink config for setting up the connection only has {application: 'trame'}. </unquote> Do I need to do this even if I want my frontend application to be hosted on a different server as the Trame server? Instead of a react app I am looking to create a vanilla js app. For this MVP we would like to use Webflow as the frontend with JS code that connects to the backend. The reason for this is simply since we have in-house Webflow frontend experience instead of React / Vue Thank you. Regards, Jeetu

…

On Wed, 15 May, 2024, 01:26 Sebastien Jourdain (Kitware), < ***@***.***> wrote: But you can look at that one <https://kitware.github.io/paraviewweb/docs/multi_user_setup.html> along with a couple of others on that same website. — Reply to this email directly, view it on GitHub <#523 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNJK77YBPVAGQ3IYG3LZCJT5ZAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIMZYGAZDQ> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jourdain]

If you want to host your react/plain-js app somewhere else, it will be more complicated and will require more knowledge of the setup which will be very hard/impossible to go over in a discussion like that. But if you bundle it with docker (for now as MVP), it will be simple.

But since you have the pieces of info in docs and code examples. You should be able to do your own integration the way you aim to.

[topcat]

If you want to host your react/plain-js app somewhere else, it will be

more complicated and will require more knowledge of the setup which will be very hard/impossible to go over in a discussion like that. But if you bundle it with docker (for now as MVP), it will be simple. Thank you so much Sebartien for the advice and help. I am attempting to package the react app found in the vtk-web-solutions example into my docker container like you suggested. I am running into the following error as seen in the console log : index-e639b0b0.js:44 POST http://localhost:8080/paraview/ 400 (Bad Request) I guess I need to understand the architecture and the setup documents and configs in the link provided by you to understand all of this a lot better. Thank you again for everything. Regards, Jeetu

…

On Wed, May 15, 2024 at 7:54 PM Sebastien Jourdain (Kitware) < ***@***.***> wrote: If you want to host your react/plain-js app somewhere else, it will be more complicated and will require more knowledge of the setup which will be very hard/impossible to go over in a discussion like that. But if you bundle it with docker (for now as MVP), it will be simple. But since you have the pieces of info in docs and code examples. You should be able to do your own integration the way you aim to. — Reply to this email directly, view it on GitHub <#523 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNJ5GN6W7C5IJ3JYQFTZCNV2LAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TINBWG42TO> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jourdain]

If you are using the trame based image, that POST http://localhost:8080/paraview/ should have worked if you changed the connect call to have the { application: 'trame' } as config. Assuming your trame app is also properly configured within docker.

[topcat]

Yes I have changed my connect call to { application: 'trame' } however that error persists. I also see the following in my terminal where I run ./scrips/run_image.sh ERROR: Launch request is missing application key.

…

On Wed, May 15, 2024 at 11:52 PM Sebastien Jourdain (Kitware) < ***@***.***> wrote: If you are using the trame based image, that POST http://localhost:8080/paraview/ should have worked if you changed the connect call to have the { application: 'trame' }. Assuming your trame app is also properly configured within docker. — Reply to this email directly, view it on GitHub <#523 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNKTIJNHMMUGHECM3QTZCORVTAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TINBZGQYDK> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jourdain]

Launch request is missing application key. See my comment on the config which has an application key...

Without code example showing what you did compare to what I told you to do, I can't help except by just repeating myself.

[topcat]

I understand Sebastien and I am sorry for the trouble. I shall read the docs closely and attempt to debug some more this looking at the link you gave earlier and code / config samples therein. Failing that I shall put up the code sample and share it with you and seek your help further. Thank you for your time and I appreciate your help. Regards, Jeetu

…

On Thu, 16 May, 2024, 01:04 Sebastien Jourdain (Kitware), < ***@***.***> wrote: Launch request is missing application key. See my comment on the config which has an application key... Without code example showing what you did compare to what I told you to do, I can't help except by just repeating myself. — Reply to this email directly, view it on GitHub <#523 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNOSANV6WUJ6U6W5C7DZCO2CVAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TINBZHEZTI> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[topcat]

Hello,

I have managed to get working the sample application with vanilla js frontend and a trame server all bundled together in a docker container, as guided to by Sebastien (thank you :) ). The problem was indeed my code and how I was connecting to the server. Once I corrected this and used the config with just {application: 'trame'} as suggested by Sebastien things are working

My next step is to have my JS client served from a different server as the trame server. This is because we would like to have a Webflow + Vanilla JS code hosted on Webflow's SaaS.

From what I have understood (and I could be wrong) this would mean, at the least, having the wsclient / SmartConnect to establish a websocket connection to the trame server in the backend.

I am still attempting to go through the code and documentation to understand how to do it. I shall keep posting here for the sake of anyone else who follows through and needs it.

The more I explore Trame the more I love it :)

Thank you so much once again Sebastien for all of your help and support.

Regards,
Jeetu

[jourdain]

Now that you have that part working when the JS execute from another host, you should be able to use a config like

{
  application: "trame", 
  sessionManagerURL: "https://trame-server/paraview",
}

Then you may need to add a configDecorator to properly process the USE_HOST placeholder so it does not pickup your Webflow server but the trame-server instead.

[topcat]

Thank you so much once again Sebastien :)

I shall try this out and post back here :) Appreciate all of your help :)

Regards,
Jeetu

[topcat]

Hi Sebastien, I am facing a problem where I keep getting a 404 not found error when the client attempts to connect to the trame server as part of the sessionManagerURL key. I first deployed my Trame server as a container onto AWS lightsail and my JS client as a S3 static site and was seeing the following errors : POST http://[AWS-trame-server-host]/paraview 404 (Not Found) WebSocket connection to 'ws://[S3-frontend-client-host]/ws' failed: * I realize that the ws URL is taking the [S3-frontend-client-host] host instead of the remote [AWS-trame-server-host]. You had suggested I use a configDecorator for this but I am not sure what I should be doing. The configDecorator for the react-client code as part of the simple-server example of vtk-web-solutions project ( https://github.com/Kitware/vtk-web-solutions/) checks if the sessionURL is defined but for us from what I understand the sessionManagerURL will return the sessionURL. Anyway, I decided to first tackle the 404 error on the /paraview endpoint thinking that this may get the correct sessionURL or at least I would then progress. I figured the above maybe because AWS lightsail was redirecting http traffic to https and the trame server wasn't configured for https and this maybe causing an issue. I wasn't sure but to test this theory out I went on to deploy two docker containers on my local system. One with the trame server and the other with the client. Both containers can ping into each other and can also get http content from each other. The config for my frontend now has the following : const config = { application: 'trame', sessionManagerURL: 'http://172.17.0.2/paraview', }; I still get the 404 not found error on the /paraview endpoint. If I log into the container with the trame server and attempt to access http://localhost:9000/paraview, I get a 404. I am sure it's something in my understanding, would appreciate if you could shed light on how I can resolve this. Thank you. Regards, Jeetu

…

On Fri, May 17, 2024 at 8:59 PM Sebastien Jourdain (Kitware) < ***@***.***> wrote: Now that you have that part working when the JS execute from another host, you should be able to use a config like { application: "trame", sessionManagerURL: "https://trame-server/paraview", } Then you may need to add a configDecorator to properly process the USE_HOST placeholder so it does not pickup your Webflow server but the trame-server instead. — Reply to this email directly, view it on GitHub <#523 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNOASV346EEVZUG37KDZCYO47AVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TINZSGI3DE> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jourdain]

Add a / at the end of paraview.

…

On Sun, May 19, 2024 at 14:29 Jeetu Golani ***@***.***> wrote: Hi Sebastien, I am facing a problem where I keep getting a 404 not found error when the client attempts to connect to the trame server as part of the sessionManagerURL key. I first deployed my Trame server as a container onto AWS lightsail and my JS client as a S3 static site and was seeing the following errors : POST http://[AWS-trame-server-host]/paraview 404 (Not Found) WebSocket connection to 'ws://[S3-frontend-client-host]/ws' failed: * I realize that the ws URL is taking the [S3-frontend-client-host] host instead of the remote [AWS-trame-server-host]. You had suggested I use a configDecorator for this but I am not sure what I should be doing. The configDecorator for the react-client code as part of the simple-server example of vtk-web-solutions project ( https://github.com/Kitware/vtk-web-solutions/) checks if the sessionURL is defined but for us from what I understand the sessionManagerURL will return the sessionURL. Anyway, I decided to first tackle the 404 error on the /paraview endpoint thinking that this may get the correct sessionURL or at least I would then progress. I figured the above maybe because AWS lightsail was redirecting http traffic to https and the trame server wasn't configured for https and this maybe causing an issue. I wasn't sure but to test this theory out I went on to deploy two docker containers on my local system. One with the trame server and the other with the client. Both containers can ping into each other and can also get http content from each other. The config for my frontend now has the following : const config = { application: 'trame', sessionManagerURL: 'http://172.17.0.2/paraview', }; I still get the 404 not found error on the /paraview endpoint. If I log into the container with the trame server and attempt to access http://localhost:9000/paraview, I get a 404. I am sure it's something in my understanding, would appreciate if you could shed light on how I can resolve this. Thank you. Regards, Jeetu On Fri, May 17, 2024 at 8:59 PM Sebastien Jourdain (Kitware) < ***@***.***> wrote: > Now that you have that part working when the JS execute from another host, > you should be able to use a config like > > { > application: "trame", > sessionManagerURL: "https://trame-server/paraview", > } > > Then you may need to add a configDecorator to properly process the > USE_HOST placeholder so it does not pickup your Webflow server but the > trame-server instead. > > — > Reply to this email directly, view it on GitHub > < #523 (reply in thread)>, > or unsubscribe > < https://github.com/notifications/unsubscribe-auth/AAATXNOASV346EEVZUG37KDZCYO47AVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TINZSGI3DE> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> > — Reply to this email directly, view it on GitHub <#523 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACH45WQRRDWPX4HESN7QKLZDEDTJAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIOJQGE3TM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[topcat]

And it worked :) thank you so much Sebastien. For the sake of anyone else who comes across this thread, once I added a / and the end of paraview and moved past this 404, I received back the sessionURL and seeing this I was able to understand how to have a configDecorator so that the USE_HOST get's replaced with the remote hostname instead of the server on which the JS client is hosted. This is exactly what Sebastien has guided me towards previously in this thread. So it works now but since my Trame server is hosted headless on AWS Lightsail, I am using vtk-osmesa and therefore the performance isn't great even for the simple cone example. AWS Lightsail also does not have a GPU instance. So my next step will be to host all of this on an EC2 instance with GPU. I remember reading somewhere that I will need to use VTK with EGL. Will need to see from where I can get this version of VTK. And then I think I will need to configure docker to access the GPU appropriately. Not sure if anything else is needed. I will post here for the sake of anyone else who stumbles upon this. Once again Sebastien, I want to thank you. I see you posting to pretty much every query on this forum and that makes this forum such a wealth of information. It's really fantastic. Sincerely appreciate it :) Regards, Jeetu On Mon, May 20, 2024 at 2:16 AM Sebastien Jourdain (Kitware) < ***@***.***> wrote:

…

Add a / at the end of paraview. On Sun, May 19, 2024 at 14:29 Jeetu Golani ***@***.***> wrote: > Hi Sebastien, > > I am facing a problem where I keep getting a 404 not found error when the > client attempts to connect to the trame server as part of the > sessionManagerURL key. > > I first deployed my Trame server as a container onto AWS lightsail and my > JS client as a S3 static site and was seeing the following errors : > > POST http://[AWS-trame-server-host]/paraview 404 (Not Found) > WebSocket connection to 'ws://[S3-frontend-client-host]/ws' failed: > > * I realize that the ws URL is taking the [S3-frontend-client-host] host > instead of the remote [AWS-trame-server-host]. You had suggested I use a > configDecorator for this but I am not sure what I should be doing. The > configDecorator for the react-client code as part of the simple-server > example of vtk-web-solutions project ( > https://github.com/Kitware/vtk-web-solutions/) checks if the sessionURL > is > defined but for us from what I understand the sessionManagerURL will > return > the sessionURL. > > Anyway, I decided to first tackle the 404 error on the /paraview endpoint > thinking that this may get the correct sessionURL or at least I would then > progress. > > I figured the above maybe because AWS lightsail was redirecting http > traffic to https and the trame server wasn't configured for https and this > maybe causing an issue. I wasn't sure but to test this theory out I went > on > to deploy two docker containers on my local system. One with the trame > server and the other with the client. Both containers can ping into each > other and can also get http content from each other. > > The config for my frontend now has the following : > > const config = { > application: 'trame', > sessionManagerURL: 'http://172.17.0.2/paraview', > }; > > I still get the 404 not found error on the /paraview endpoint. > > If I log into the container with the trame server and attempt to access > http://localhost:9000/paraview, I get a 404. > > I am sure it's something in my understanding, would appreciate if you > could > shed light on how I can resolve this. > > Thank you. > > Regards, > Jeetu > > On Fri, May 17, 2024 at 8:59 PM Sebastien Jourdain (Kitware) < > ***@***.***> wrote: > > > Now that you have that part working when the JS execute from another > host, > > you should be able to use a config like > > > > { > > application: "trame", > > sessionManagerURL: "https://trame-server/paraview", > > } > > > > Then you may need to add a configDecorator to properly process the > > USE_HOST placeholder so it does not pickup your Webflow server but the > > trame-server instead. > > > > — > > Reply to this email directly, view it on GitHub > > < > #523 (reply in thread)>, > > > or unsubscribe > > < > https://github.com/notifications/unsubscribe-auth/AAATXNOASV346EEVZUG37KDZCYO47AVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TINZSGI3DE> > > > . > > You are receiving this because you authored the thread.Message ID: > > ***@***.***> > > > > — > Reply to this email directly, view it on GitHub > < #523 (comment)>, > or unsubscribe > < https://github.com/notifications/unsubscribe-auth/AACH45WQRRDWPX4HESN7QKLZDEDTJAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIOJQGE3TM> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> > — Reply to this email directly, view it on GitHub <#523 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNIBVEJKJH4SYYVEWMTZDEFSJAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIOJQGI2DK> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jourdain]

For vtk-egl see this requirements but replace vtk-osmesa by vtk-egl.

Also make sure you do not have a regular vtk previously installed.

pip uninstall vtk vtk-osmesa
pip install vtk-egl --extra-index-url https://wheels.vtk.org

[topcat]

Thank you so much Sebastien :) will give this a go. Appreciate it :) Regards, Jeetu

…

On Mon, 20 May, 2024, 21:00 Sebastien Jourdain (Kitware), < ***@***.***> wrote: For vtk-egl see this requirements <https://github.com/Kitware/trame/blob/master/examples/deploy/docker/VtkRendering/setup/requirements.txt> but replace vtk-osmesa by vtk-egl. Also make sure you do not have a regular vtk previously installed. pip uninstall vtk vtk-osmesa pip install vtk-egl --extra-index-url https://wheels.vtk.org — Reply to this email directly, view it on GitHub <#523 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNJRJYRB4BEW24OH5FDZDIJK7AVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIOJYGAZDS> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[topcat]

Hello, And I now have the Trame server for my application working in a docker container running on a AWS EC2 instance with GPU (gs4dn.xlarge instance type) based on the guidance given by Sebastien (thank you :) ). I have EGL working in the container. My frontend is being hosted separately from the trame server and remote rendering works really well :) For the sake of anyone who follows this thread, I would like to point out what I needed to do to get EGL working in the container and able to access the GPU. The section on EGL caution at https://kitware.github.io/trame/guide/deployment/docker.html came in very handy to me. However, I needed to include the -e NVIDIA_DRIVER_CAPABILITIES=all, without which EGL wasn't fully accessing the GPU and I was seeing a slight unresponsiveness. Therefore, the command in the document is : docker run --gpus all -it nvidia/opengl:1.0-glvnd-runtime What I needed to do was : docker run --gpus all -e NVIDIA_DRIVER_CAPABILITIES=all -it nvidia/opengl:1.0-glvnd-runtime Now, currently I have Apache in the trame container listening on port 80 (http). I would the trame container to work with https / wss. I came across this thread on the forum that discusses configuring trame / apache for https / wss - #334 The one thing I am not sure of there is the apache config the OP has mentioned in his final solution still has <VirtualHost *:80> shouldn't this be 443? Once again Sebastien, thank you. Trame is beautiful and you have made this journey of learning it so much more pleasurable. Appreciate all of your help :) Regards, Jeetu On Mon, May 20, 2024 at 9:02 PM ***@***.*** < ***@***.***> wrote:

…

Thank you so much Sebastien :) will give this a go. Appreciate it :) Regards, Jeetu On Mon, 20 May, 2024, 21:00 Sebastien Jourdain (Kitware), < ***@***.***> wrote: > For vtk-egl see this requirements > <https://github.com/Kitware/trame/blob/master/examples/deploy/docker/VtkRendering/setup/requirements.txt> > but replace vtk-osmesa by vtk-egl. > > Also make sure you do not have a regular vtk previously installed. > > pip uninstall vtk vtk-osmesa > pip install vtk-egl --extra-index-url https://wheels.vtk.org > > — > Reply to this email directly, view it on GitHub > <#523 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AAATXNJRJYRB4BEW24OH5FDZDIJK7AVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TIOJYGAZDS> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

[jourdain]

Correct it should be 443. And on the 80 you could setup a redirect to 443.

[topcat]

For the sake of anyone who stumbles across this thread, I managed to configure the trame docker container to work with a self-signed SSL certificate. I have managed to deploy this on a EC2 instance with GPU (gs4dn.xlarge instance type) and configured my instance and docker so the GPU is available inside the docker and works with EGL. I am using vtk-egl. On the frontend I have got a vanilla JS client that talks to our trame server using the wslink library. Should someone have an interest in this then please ping me and I can probably document these steps and create a sample. Thank you very much Sebastien for all of your help and your patient guidance. Appreciate it :) Regards, Jeetu

…

On Sun, May 26, 2024 at 8:45 PM Sebastien Jourdain (Kitware) < ***@***.***> wrote: Correct it should be 443. And on the 80 you could setup a redirect to 443. — Reply to this email directly, view it on GitHub <#523 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAATXNI3DGTVOK76X4BEXZTZEH4BPAVCNFSM6AAAAABHWYRQKSVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TKNRSGEZDM> . You are receiving this because you authored the thread.Message ID: ***@***.***>