I noticed that on Ubuntu 20.04, a lot of software is made available through snap and it does not even appear in the distribution's "sources".
This is all good, but when you try to run an application that has graphic output and have it connect via remote X11 to your machine, snap seems to be Unable to open X display. Which is strange, because gnumeric (for example) runs as a remote X11 application just fine.
I suspect that this is a configuration problem. Any ideas on how to run a remote X11 application via snap appropriately?
Try running this on the remote host before launching the snap application:
export XAUTHORITY=$HOME/.Xauthority
Related
I'm trying to debug some code running on a server using the Visual Studio 2015 Remote Debugger. I can get connected and view all the running processes but when I try to connect to the w3wp.exe process it is having none of it.
From my fiddling I've noticed that it will connect to apps that are running in the Managed (v4.0.30319), x86 mode, however if it's just x86 it doesn't connect and I'm guessing this is because it's been compiled and isn't in MSIL.
I've copied over the files from my machine, by passing all the precompilation that we were doing, as much config set to debug as I can find and even still when I try to attach it still reports x86 with no mention of any managed code.
Just to clarify, this is what I see when attaching on my local machine:
And this is the same app running on the remote server:
Does anyone know how to force the server to run in managed mode, or is there another reason I can't connect to the remove debugger?
Thanks
maybe the question title is a little bit weird... I will explain more here...
Basically, suppose that I have a local laptop and a remote machine. I can of course run R tasks on the remote machine itself. Now I wonder if I can connect the remote machine from my local R and run the tasks on the remote machine while I still call these things in my local machine... that is I don't need to open R or type/source R codes on the remote machine.
One related thing is that I am considering to develop a small app with R-shiny. As far as I understand, though I just started looking into it, one can have R installed the local machine and use the app - the calculation will be done in R on the same machine. This question is essentially that if it is possible to add a feature of “connection to a calculation back end” in the app, i.e. by connecting to the remote machine, the task is done on the remote machine when one uses the app on his local machine.
You can do this with RStudio Server. It runs RStudio inside a web server on the remote machine. You connect to it from the web browser on your local laptop, so you can still enter your R code on the local machine.
RStudio Server can be downloaded here:
https://www.rstudio.com/products/rstudio/download-server/
I am aware that I can use RStudio server on my remote machine, and connect to that via a web browser on my client machine.
However, I would prefer not to use my web browser. Is it possible to set up RStudio Desktop to connect to the RStudio Server?
Many thanks
There is an alternative: create Rstudio Server webapp.
This is how I did it:
Login to the rstudio server from the browser using serverip:8787
On the menu bar, go to File (or Tools in some cases) -> Create application shortcut. This dialogue box will appear:
Rstudio shortcut
Add it to your desktop by selecting and run it as if you're running rstudio desktop version.
Rstudio webapp screenshot
Hi I'm trying to make my first meteor app. I just made an app and have run it. I've also upgraded to latest meteorite and meteor. I just created a new app and have run it.
App running at: http://localhost:3000/
But I cannot view this webpage on my host computer. I'm running meteor on an ubuntu vm.
It was working before I updated meteorite and meteor and installed the iron-router package to an app I'm working on.
check in the console you might have the error as template is not defined, Meteor is not defined etc...
I too was unable to connect to the Meteor App that was being developed on a centos VM.
The followuing worked:
Setup:
Windows 7/8.1 with Following tools installed:
Putty
VMW Workstation
Centos VM (Server or GUI based)
Chrome/Mozilla Browsers
Launch the VM from VMW Workstation
Console into the VM using Putty
Launch the Meteor App. Will show its listening at localhost?3000
Launch a new session of Putty.
Go To SSH-Tunnels - Source: 3000, Destination - Localhost:3000. Save this settings
Connect to the VM with these settings
Launch the browser and and navigate to localhost:3000
Worked for me, hope does for all
As you have determined, the message on the vm regarding localhost:3000 refers to the vm that is running meteor, and localhost on your computer refers to your computer.
The secure fix for this is to create an encrypted tunnel to connect localhost:3000 on your computer to localhost:3000 on the remote computer.
Start the meteor app on the VM so that you get the message about it being ready on localhost:3000
On the local computer open another terminal window and initiate a second connection to the VM with:
ssh -L 3000:localhost:3000 yourUSER#remoteHOST
This assumes you are running Linux. If you are on Windows, look at the options for your SSH client. It may have similar options to create tunnels.
Open a web browser on your local computer and go to http://localhost:3000
The ssh tunnelling software will sense the connection to localhost:3000 on your computer and will connect you to localhost:3000 on the remote, forwarding the data through an encrypted tunnel.
If this seems like a lot of trouble, there are paid developer platforms like http://nitrous.io that can run meteor and have a web based IDE that can simplify this sort of thing for you so you do not need to run the tunnel. Another way to simplify is to not use a remote VM, but install Meteor on the home computer and only copy the code to a VM when it is finished and ready for production.
If you don't want to run on localhost:3000 at all, but on the webserver on port 80, you might check to see if there is an environment variable that switches the code from development mode to production.
I am transitioning my debian setup into one where all debian-repository external apps run in dedicated docker containers.
In this context rstudio, of which I am a heavy user, has me puzzled ... does anybody have insight into whether it's possible to run it as a client to remote R installation?
What is a very cool feature of RStudio is RStudio Server. You install RStudio Server on you Ubuntu server and log in to a specific port where RStudio Server is running. You then get your full RStudio interface in your web browser. This allows you to run all your R analyses from any computer that has a modern browser and an internet connection.
R then runs on the remote server, asking almost no resource from the computer you are connecting from.