First of all, I think this is more of a Linux issue as the problem seems to be on a linux-flavoured Docker container, but I'm happy to accept that I can do something to the team city config to overcome this.
I'm also not very experienced with Linux, Docker or node/npm, though I do have a lot of development experience and am very comfortable with command line interfaces in general.
Background
We currently have Team City set up as a build server, for building a variety of projects:
.Net Framework,
.Net Core
Angular CLI
A couple of simple websites which use node packages to generate HTML from Markdown.
The server is running as a Docker container using Docker for Windows on a Windows Server box, and this is working well.
We have one Windows 10 Build agent (a VM) which is also working fine, and builds all the .Net and .Net Core stuff fine.
The simple docs site stuff primarily uses the markdown-to-html node package, so its build steps simply get all the source .md files and compile to html with markdown-to-html, plus use some other npm packages for SASS compilation and minification of js etc. No actual node code as such, just some jQuery. In order to not tie up the other agent, and because this stuff can run happily on Linux, I want to have this running on a small docker image rather than a full VM build agent somewhere.
I previously successfully used a node.js team city agent docker image (either jacobpeddk/teamcity-agent-nodejs or omez/teamcity-agent-nodejs - can't recall) which did work for a time, though I had issues with being able to install some npm packages globally in build scripts, which meant I had to get a bash terminal into the container and run some manual npm commands. I also I think had to run apt-get install zip to get a zipping step to work. This worked fine for a while (weeks).
I added some extra JS stuff to one of these simple projects, and suddenly I was getting errors when trying to build. I (perhaps stupidly) decided that this was probably due to the container having older versions of node and/or npm etc, so I attempted to update this by getting a bash shell into the container, installing nvm and updating node.js & npm.
This ended up with a rather broken container (node errors), so I thought I'd instead start again, but actually start with the jetbrains/minimal-build-agent Docker image instead, with the aim of ending up with a nice bespoke image for our needs specifically (as I couldn't find a very up-to-date pre-existing one)
I've running a Bash shell directly on the build agent container by executing this on the host:
docker exec -it basicagent /bin/bash
then from there I've installed nvm, Python (required for node install step) and node:
curl -o- https://raw.githubusercontent.com/creationix/nvm/v0.33.2/install.sh | bash
export NVM_DIR="$HOME/.nvm"
[ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh" # This loads nvm
[ -s "$NVM_DIR/bash_completion" ] && \. "$NVM_DIR/bash_completion" # This loads nvm bash_completion
apt-get update
apt-get install python 3.6
nvm install v8.11.1 (matching version on my dev machine)
npm install -g markdown-folder-to-html (npm package I previously found I had to install globally)
apt-get install zip (just used for a build step to zip up artifacts)
If I now run (via the bash shell) npm -version I get back 5.6.
If I try to get a build to run that uses npm in a command line step, then I get this error in the build log:
/opt/buildagent/temp/agentTmp/custom_script2764770419520852926: npm: not found
I wondered if it was an issue with the user/path that the team city agent process is using vs. the one I'm using in Bash, so I added the following to the build script:
echo PATH = $PATH
echo user var = $USER
echo user via 'id':
id -u -n
the output of which is:
PATH = /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
user var =
user via id:
root
So it's running the agent as root, and doesn't appear to have node in the $PATH at all.
If I run the above directly from Bash however, I can see that I am root, but my $PATH is different:
PATH = /root/.nvm/versions/node/v8.11.1/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
root
So I'm now confused: I've re-started the container and this has had no effect - it seems that when I'm logged in as root manually I have a certain path set, but when the build agent service is running as root it's different.
I have no idea why this happens, but I've basically worked around the problem by adding:
export PATH=$PATH:/root/.nvm/versions/node/v8.11.1/bin
to the top of every build step that uses npm in a script. To my mind this seems a rather daft thing to have to do - considering this used to work without this, and the only real difference is possibly a slightly different flavour of linux container. AFAIK the original build agent container was based on the jetbrains minimal-build-agent one, so unless they've changed what they base that on it should be roughly the same...
I also had to change the compressor being used in a node-minify build step from gcc (google closure compiler) to babel-minify as the former was basically hanging indefinitely, but that is a separate problem (though also something that was fine and now isn't...)
Thanks to anyone who took the time to read... though I do wonder if one-day I'll exhaust my own research options, and finally go ask the internet and actually get someone respond - for some reason whenever I get to the point where I have to ask, it always seems no-one else has the answer either and I end up having to work it out myself. It's probably character-building though I suppose.. (this isn't just SO - I've found this be the case for over 15 years on various forums about various things...)
Please tell me how i can solve it. Iam using window 7 32 bit:
By default, Visual Studio Code runs shell commands like npm in a loginless shell. If you installed NVM, Visual Studio Code may have no indication where to find npm to run it.
Put the following lines are in .bash_profile: (Note: Not .bashrc.)
export NVM_DIR="$HOME/.nvm"
[ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh" # This loads nvm
[ -s "$NVM_DIR/bash_completion" ] && \. "$NVM_DIR/bash_completion" # This loads nvm bash_completion
Add this setting to settings.json:
"terminal.integrated.shellArgs.linux": ["-l"]
On Mac OS , I switched to the zsh and got this fixed. These following steps fixed the issue :
On your VsCode in Mac : shift + command + P .
On the Prompt > type : Terminal: Select Default Profile , then "Click it". Note, as you type you will find this option in the auto-complete .
Click the option for zsh or your desired shell.
Restart VSCode .
now npm will work in your vscode terminal.
I too had this issue.
To solve this follow the below steps.
Make sure you have npm installed
- go to command prompt & type npm -v
- if a version comes out it's installed, else go to https://nodejs.org/en/ and download same.
Then come to vs code and set deafult shell to cmd.
to do so,
- Press Ctrl+Shift+P and type Select Default Shell
- Select Command Prompt.
- Press Ctrl+`
- Type npm -v and see npm works. :)
I got this error after having just installed nodejs/npm, and the way i resolved, while still being able to use git bash terminal in vscode was to simply restart my computer (I hadn't done this after node installation).
I needed to kill the terminal in VS Code and restart a new terminal to get npm to show as installed.
As you're aware, you can configure your environment variables in windows at 2 levels.
At user level
At system level
When VS code is started, it picks the variables in path at a user level and not at system level because you haven't run the application as Administrator.
Just copy-paste your node path from System Variables to User Variables
This does the trick for windows.
Thanks.
If you just installed your node while your VScode was running. Restart your vscode it should start working.
You need to install npm first, https://www.npmjs.com/get-npm and make sure npm command is accessible using terminal/command prompt.
You can also use https://marketplace.visualstudio.com/items?itemName=eg2.vscode-npm-script This extension supports running npm scripts defined in the package.json file and validating the installed modules against the dependencies defined in the package.json.
On mac, I switched the default shell from bash to zsh and it fixed the issue.
I'm gonna add an answer just for help others because this question is 2 years old.
If you can run the npm start (NPM SCRIPTS below the Explorer view) but you are not able to launch the command with a shortcut: Then check npm extension is installed and/or enabled for the current workspace/folder.
Otherwise vsCode will not be able to run the script and you're going to get a message in a little box like this
command 'npm-script.start' not found
Maybe the NPM Scripts View below the explorer is not available without the plugin I'm not sure
For me, restarting VS code and even my PC didn't work, but after restarting VS code via "reload" (not restart, not refresh) from ctrl+shift+P, then in the top right of the terminal clicking the leftmost button and clicking "kill terminal" then restarting VS code it worked.
I had this issue too and I'm running QubesOS which uses virtualization.
I could manage to get it running until I'd a standalone VM.
That solved my problem.
Just mentioning for the ppl using virtualization like virtual box or VMware.
Is it possible for VS Code to use node version specified by NVM?
I have 6.9.2 installed locally. Even after switching to another version, from the OS X terminal (not the VS Code terminal), restarting VS Code, VS Code still shows using 6.9.2.
OS X terminal
MacBook-Pro-3:~ mac$ node -v
v7.8.0
VS Code Terminal
MacBook-Pro-3:QB-Invoice-API mac$ node -v
v6.9.2
In VS Code:
go to your launch.json file
add the runtimeVersion attribute inside configurations as shown below
In this example, we are assuming 4.8.7 is already installed using nvm:
{
"version": "<some-version>",
"configurations": [
{
"type": "node",
"runtimeVersion": "4.8.7", // If i need to run node 4.8.7
"request": "launch",
"name": "Launch",
"program": "${workspaceFolder}/sample.js"
}
]}
The solution is to set alias default. In the OS terminal run -
nvm alias default 7.8.0
Open vscode, now running node -v returns 7.8.0
It seems vscode takes up this (alias default) value and not the node version that is set by nvm use X.X.X
Restart VS code for it to pick up the changes.
add runtimeExecutable to your .vscode/launch.json like this
{
"type": "node",
"request": "launch",
"name": "App",
"program": "${workspaceRoot}/index.js",
"runtimeExecutable": "${env:HOME}/.nvm/versions/node/v6.9.2/bin/node"
}
I had the same problem of being unable to keep my node version specified trough nvm in my OS X environment not only with VSCode but also with Atom Editor (using the platformio-ide-terminal package for managing the integrated terminal in it). None of the suggestions in the previous answers worked for me, besides me not using the debugger but using gulp and grunt for specific tasks. Apparently nvm does not get along with the integrated terminals or sub shells at least in these editors because when loading them the environment variable $PATH is modified internally and does the following according to a comment by one of the contributors of this package in this issue reported here NVM fails to load within nested shell #1652:
" #charsleysa I know why nvm is throwing this error. In your subshell, somehow the /usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin part of your PATH has been moved from the end of the PATH to the start.
When nvm is then started, it calls nvm_change_path (my contribution changed it to this from nvm_prepend_path), which modifies the nvm-relevant part of the path in place.
Nvm then checks the current npm prefix by asking npm what it is. Since /usr/local/bin/npm now has precendence, it reports /usr/local/bin.
Nvm then checks whether the current prefix as reported by npm is in the directory tree of the current nvm node version (at this stage, the installation directory of the node version your default nvm alias resolves to).
The prefix is not part of that tree, so it deactivates itself (calling nvm_strip_path in the process, which is why there's no nvm-related path in your subshell's PATH), and bails with the error you're getting.
macOS's /etc/profile (or /etc/zprofile) calls /usr/libexec/path_helper, which does the PATH switcheroo.
In the parent shell, the PATH doesn't yet have an nvm dir in it, so by the time nvm runs, it prepends its directory to the path. But in the subshell, PATH has been reconfigured by macOS to put any non-system directories at the end and we have the problem."
I was always getting this message when launching any integrated terminal:
nvm is not compatible with the npm config "prefix" option: currently set to "/usr/local"
Run npm config delete prefix or nvm use --delete-prefix vx.x.x --silent to unset it.
What I did to solve this in my case was the "workaround" part of that same issue reported which is essentially the following:
Reset the path by adding the following line inside my ~/.bash_profile at the very top before anything else:
PATH="/usr/local/bin:$(getconf PATH)"
And after that no more warnings when I launch any integrated terminal on both editors and I can interact with nvm to switch between any node version easily and without problems at all.
Here it is another alternative just in case this one doesn`t help that much.
Some of the answers provided are correct and upvoted, but somewhat incomplete. This procedure worked for me:
Open the terminal window inside VS Code and run node -v. You'll get for instance v10.12.0.
Open a terminal window outside VS Code Change your node version with nvm (ie. nvm use v12.14.0)
Cmd+ Shift + p and choose Preferences > Open Settings (JSON)
Add "terminal.integrated.shellArgs.osx": [] to your user config
Cmd+ Shift + p and choose Shell Command: Install 'code' command in PATH
Close VS Code.
Open a terminal window and run code. This will open VS Code with a new and updated bash / zsh session.
Open the terminal window inside VS Code and run node -v. You'll get v12.14.0.
Bonus: If you always want to get a particular node version on VS Code's terminal, set it as default by opening a terminal window outside VS Code and running:
nvm alias default v12.14.0
I am using oh-my-zsh and it too was not using the node version specified by nvm. Tried several suggestions posted here, but the only way I managed to resolve this issue was by adding the following line to the top of ~/.zshrc
PATH="/usr/local/bin:$(getconf PATH)"
I had the same problem, but the above answers didn't help.
Apparently the default shellArgs for osx are set to bash while I'm using zsh. I solved the problem by setting the shellArgs in my user settings to an empty array:
"terminal.integrated.shellArgs.osx": []
A alternative solution I've found is to simply launch code from the shell after you pick your node using nvm.
You need to first open the command pallet and select "install 'code' into
path".
And then launch a terminal and select your node via nvm and then launch "code".
I had this same issue and I found a strange workaround that may be helpful to someone else in the future.
If I do not set eslint.runtime my system was running node v10.11.0 for eslint server, whereas I wanted it to be running v12.13.0 which I had installed and made default via nvm.
I found that the v10 version of node was installed by brew based on #franziga's answer but my desired version of node was installed by nvm. So, I uninstalled v10.11.0 via brew and closed/reopened VS Code. Strangely, eslint was still reporting that it was started using v10.
I tried running a shell without any changes to my PATH in any startup scripts, and the version of node was still correctly pointed to v12 as expected, but VS code still starts up v10 for eslint.
I'm not sure how to check the path of the executable that is being run by eslint, and if I open an integrated terminal everything works fine with the expected version of node (v12).
Solution (for me):
I found that if I set "eslint.runtime": "node" in settings.json that it will now use whatever version of node was active when I opened vscode using code . on the terminal. Just "node" - no path.
Lots of complicated answers here. In my case, this was caused by node having previously having been installed. Fixed it by deleting the following directories: rm -rf /usr/local/bin/npm and rm -rf /usr/local/bin/node, then running nvm use default in VSC to pick up the node version installed by nvm.
Particularly with the shell I had no problems, but you may:
check your shell is properly configure or change it (you might be using different shells for vscode or your terminal)
check your env and if it's not properly set use the terminal.integrated.env.<platform>
I had issues with vscode itself and no solution could help me. So I finished using the following launch script.
{
"type": "node",
"request": "launch",
"name": "Launch Program",
"program": "${workspaceFolder}/server.js",
"runtimeExecutable": "/bin/bash",
"runtimeArgs": ["-c", ". ~/.nvm/nvm.sh;nvm run default \"$#\"", "dummy"]
},
this assumes you have it configure for bash (otherwise change it to your shell) and you want to use the default node version as configured by nvm (you may also change it).
Note: The "dummy" parameter is required so the rest of the parameters are properly parsed.
A longer explanation of "dummy": Shell scripts use positional parameters where the first one will be the script location itself (addressed by $0), when using the -c flag the script is read inplace an there is no $0 being set. vscode will pass some arguments, like the node start script location which will be wrongly interpreted, so "dummy" pushes all parameters one spot. It can be just anything, but it must be there.
Setting the default alias only worked for me after closing all instances of VS Code. Simply reloading the window wouldn't work. nvm ls would show the default alias set correctly but would keep using the version set when VS code was first opened (across all windows).
There's also the issue of having multiple node versions across repos, which is really what nvm is meant to solve. In the end I added the following to the bottom of my .zshrc file:
[ -s "./.nvmrc" ] && nvm use
Essentially when a new shell starts, it checks to see if a non-empty .nvmrc exists in the current directory and if so, uses that version. If that version is not installed you will get a message saying so. After running nvm install it should load properly in all new terminal sessions.
You can also use the automatic version switching on directory changes shown in the nvm README as #asiera pointed out. Although a project terminal will typically always open in the same directory as your .nvmrc file, so this solution is a bit simpler and only runs on terminal startup.
I found that setting the node version locally in a sub shell before calling code works well, while not changing the version in the current shell, e.g.
$ (nvm use 14; code .)
Therefore, to make it work transparently for any project folder, create a file .precode in the project folder with shell commands to source before starting code - e.g.,
nvm use 14
Then add to ~/.bashrc
pre_code(){
if [ $# == 1 ] && [ -f ${1}/.precode ] ; then
echo "(source ${1}/.precode ; `which code` ${#})"
(source ${1}/.precode ; `which code` ${#})
else
`which code` ${#}
fi
}
alias code='pre_code'
(Note: Run source ~/.bashrc in any shell opened before the edit in order for the edit to take effect.)
Then, assuming the necessary file ~/myproject/.precode exists, starting code with
$ code ~/myproject
will result in some diagnostic output on the shell, e.g.
source github/myproject/.precode
Now using node v14.15.1 (npm v6.14.8)
as well launching a new vscode window with the correct node version in the terminal window and debugger. However, in the shell from which it was launched, the original node version remains, e.g.
$ node -v
v12.19.1
I tried all the suggested solutions but nothing was working.
/usr/local/bin/node was pointing to somewhere. i made a symlink to a specific nvm node folder and that was solving the issue for me:
ln -s /Users/mad/.nvm/versions/node/v11.1.0/bin/node /usr/local/bin/node
I have the same problem and I found that I have node installed by brew and nvm. I uninstalled node installed by brew and the versions on both terminal and visual studio code are the same now.
You don't need to modify your default node version. The following example assumes that node 6 is your default version and you want VSCode to reference version 7 of node:
# open a terminal instance
nvm use 7
code . # or project folder instead of "."
# when VSCode start, you may use ctrl+` to open the integrated terminal
# then check the node version in the integrated terminal
node -v # should print 7
After reading this thread and testing almost all suggestions, I found a very simple solution if you are using nvm: Add nvm use in the command.
It's gonna take a little more time to start the debugger, but it is worth it to me because now I don't have to execute nvm use and open Vscode by the terminal every time I start working on a different project.
Here is my .vscode/launch.json file. Good luck!
{
// Use IntelliSense to learn about possible attributes.
// Hover to view descriptions of existing attributes.
// For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
"version": "0.2.0",
"configurations": [
{
"command": "nvm use && yarn start",
"name": "Launch",
"request": "launch",
"type": "node-terminal",
},
]
}
I wanted the solution to be workspace specific and not requiring any action on my part (not having to redo nvm use <version> each time i start a terminal)
The solution I found:
create the .nvmrc file at the root of my project that old the node version I want to use as stated in nvm ReadMe
adding the automatic activation script in my ~/.zshrc also in the ReadMe (bashrc script also in the readme)
So, your nvm is configured well, but other version of node STILL keeps taking over?
Remove all non-nvm versions of node:
brew uninstall --force node (yarn will be fine without system node)
Other version installed from pkg or other non-nvm method
Re-login. Now, nothing can be fighting for path with nvm no matter how is shell launched.
Note: When installing/upgrading yarn, use brew install yarn --without-node
VSCode Shell args seems to be deprecated, here's update using profiles in VS Code's settings.json:
This gets rid of the -l terminal argument turning it into an interactive shell vs a login shell.
"terminal.integrated.profiles.osx": {
"zsh (normal - me)": {
"path": "zsh",
"args": []
}
},
"terminal.integrated.defaultProfile.osx": "zsh (normal - me)"
Thanks! the answers here for explanation and here for old args way:
For me I simply did:
# in a separate terminal (i.e not in VScode teminal)
nvm install <node version>
then in VScode terminal:
nvm use <the node version you installed>
In case you'd like to set the Node version for your Visual Studio Code NPM script runner, here's what works on a per-project basis. So, without having to set global nvm defaults.
By "NPM script runner" I mean the hover-and-execute-scripts functionality directly in package.json:
Step-by-step
Place an .nvmrc file containing the project's Node version into the root folder of your project.
Enable automatic environment as described here: https://github.com/nvm-sh/nvm#deeper-shell-integration.
Open VS Code's settings.json and define your preferred shell (in my case, it's zsh). For the automation profile, it's important to define a login and interactive shell (arguments -l and -i):
"terminal.integrated.automationProfile.osx": {
"path": "/bin/zsh",
"icon": "play",
"args": ["-l", "-i"],
},
"terminal.integrated.profiles.osx": {
"bash": null,
"zsh": {
"path": "/bin/zsh",
"icon": "star",
}
},
Result
Opening a new shell triggers NVM (The icons show which setting is working):
And running an NPM script triggers NVM:
Cheers!
following solution worked for me
first install and use the desired node version with nvm with these commands: nvm install v16.13.1 and nvm use v16.13.1.
then get the pathname of the currently using node with which node command on Linux.
it will be something like this /usr/local/nvm/versions/node/v16.13.1/bin/node
finally use this pathname in launch.json for runtimeExecutable option.
the launch.json file
{
"version": "0.2.0",
"configurations": [
{
"type": "pwa-node",
--> "runtimeExecutable": "/usr/local/nvm/versions/node/v16.13.1/bin/node",
"request": "launch",
"args": ["testcode/hunter.js","127.0.0.1:9229"],
"name": "Launch Program",
"skipFiles": [
"<node_internals>/**"
],
"program": "${workspaceFolder}/index.js"
}
]
}
If none of this answers worked for you,
If you have previously installed node by downloading and unzipping it.
Go to usr/local/lib and there will be this guy sitting around named nodejs.
Kick him out.
And set nvm alias default to preferred version again.
That's it, happily ever after. At least worked for me though.
According to the docs of nvm, you need to add this code snippet to your ~/.bashrc, ~/.profile, or ~/.zshrc file, so open the file and paste this in, restart vscode and enjoy nvm
export NVM_DIR="$HOME/.nvm"
[ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh" # This loads nvm
[ -s "$NVM_DIR/bash_completion" ] && \. "$NVM_DIR/bash_completion" # This loads nvm bash_completion
source: https://github.com/nvm-sh/nvm#manual-install
Did not tried all of the solution, but for me updating nvm simply worked.
Just follow the installation here and make sure that you bash_profile is updated.
None of the other solutions worked for me.
So I ran nvm alias default node and this fixed it for me.
I tried several of the options here at the top and they didn't work. I found a simple solution. In the VS Code terminal:
Click the down arrow on the terminal dropdown
Select Default Shell
Select 'bash'
Try node -v and that should return the correct version that was set as default nvm alias default v12.14.0
Check your default interactive shell on your MAC.
If it's zsh, how about setting the terminal in VS Code to zsh mode as well? Then you can use the specified node version on the Mac. This works for me.
I'm using macOS Big Sur v11.2.1 + VS Code v1.55.1
Setting pictrue
sudo rm -rf /usr/local/opt/node#<YOUR_NODE_VERSION>
then restart the Visual Code