Dozens of python processes started (after update to 2018.11.0)

[FreeHarry]

Environment data

• VS Code version: 1.29.1
• Extension version (available under the Extensions sidebar): 2018.11.0
• OS and version: Windows 10
• Python version (& distribution if applicable, e.g. Anaconda): 3.6.7
• Type of virtual environment used (N/A | venv | virtualenv | conda | ...): No
• Relevant/affected Python packages and their versions: XXX

Actual behavior

Since update to 2018.11.0 after some minutes of work dozens of python processes are started. Currently ~150 processes and 4.5 GB of memory is used!!

It seems each time I switch to another python file tab or change and save a python file a new python process is started. Very strange.

[FreeHarry]

I test it currently also on my Linux machine (Fedora 29). The same problem occurs.

[syegulalp]

I'm experiencing something similar. When opening Python files in a project, new instances of Python are started in the background but don't close unless I close and reopen the folder (using File | Close Folder).

[syegulalp]

The problem seems related to the use of jedi. When I disabled that and reverted to the MS Language Server, the problem ceased.

[DonJayamanne]

@Nutznieser

Since update to 2018.11.0 after some minutes of work dozens of python processes are started. Currently ~150 processes and 4.5 GB of memory is used!!

Please could you send a screen shot of the processes spawned along with the arguments passed to the processes.

[omouri]

The problem seems related to the use of jedi. When I disabled that and reverted to the MS Language Server, the problem ceased.

For me it was the opposite. I was seeing huge amounts of RAM being used by Microsoft.Python.LanguageServer. When I ENABLED Jedi I stopped seeing the issue.

[DonJayamanne]

For me it was the opposite. I was seeing huge amounts of RAM being used by Microsoft.Python.LanguageServer. When I ENABLED Jedi I stopped seeing the issue.

Please could you create a separate issue for this

[IVNSTN]

Ubuntu 17.10

vs code:

Version: 1.29.1
Commit: bc24f98b5f70467bc689abf41cc5550ca637088e
Date: 2018-11-15T19:07:43.495Z
Electron: 2.0.12
Chrome: 61.0.3163.100
Node.js: 8.9.3
V8: 6.1.534.41
Architecture: x64

Number of processes raises while editing code in VS Code until consumes all memory.
All of them are gone after Code shutdown.
Disabling Jedi intellisense fixes the issue but linter starts failing on every import and intellisense does not work at all.

[FreeHarry]

Here is a screen shot from the internal process explorer. Very similar to that from @IVNSTN.

[WhistleWhileYouWork]

I can also confirm this behavior. It happens during code completion. Every time you start typing, a conhost.exe process launches python.exe with the jedi script indicated in the comment above. After disabling jedi in the settings and reloading, the number of processes remain stable.

[2-5]

• Version: 1.29.1 (system setup)
• Commit: bc24f98b5f70467bc689abf41cc5550ca637088e
• Date: 2018-11-15T19:13:36.375Z
• Electron: 2.0.12
• Chrome: 61.0.3163.100
• Node.js: 8.9.3
• V8: 6.1.534.41
• Architecture: x64

Windows 10

Same problem, after working for a few hours hundreds of python Jedi processes are running, consuming in total 5 GB even if I've set "python.jediMemoryLimit": 128. These processes seem to never exit, even if I close all open Python files in Code.

Exiting VS Code causes the whole computer to freeze for 5 seconds, even the mouse is lagging, probably because all of these Python processes need to be terminated. I have a fast SSD.

The command line for the Python Jedi processes is:

...\python.exe c:\Users\USER\.vscode\extensions\ms-python.python-2018.11.0\pythonFiles\jedi\evaluate\compiled\subprocess\__main__.py c:\Users\USER\.vscode\extensions\ms-python.python-2018.11.0\pythonFiles 3.6.4

[DelBiss]

For me, it happen with python refactor.py AND jedi
When looking with the VS Code process explorer, on both case, all process had a child console-window-host (Windows internal process)

With python refactor.py the issue happen when there is a syntax error in the user file when the user try to refactor a variable.

Code example that will create this issues:

VariableToRefactor = "RefactorMe"
foo = "another"
syntaxError = "This is an error
VariableToRefactor = "Something"

VS Code process explorer:

Mem(MB)	pid	Name
57	7592	code main
54	11376	     shared-process
228	12820	     window (file.py - project - Visual Studio Code)
78	6624	         extensionHost
14	5608	             python refactor.py c:\Users\USER\workspace\project
9	11600	                 console-window-host (Windows internal process)
4	13036	             "C:\Users\USER\AppData\Local\Programs\Microsoft VS Code\Code.exe" "c:\Users\USER\AppData\Local\Programs\Microsoft VS Code\resources\app\extensions\html-language-features\server\dist\htmlServerMain" --node-ipc --clientProcessId=6624
14	14272	             python refactor.py c:\Users\USER\workspace\project
9	576	                 console-window-host (Windows internal process)
4	21084	             electron_node server.js 
14	22296	             python refactor.py c:\Users\USER\workspace\project
9	44976	                 console-window-host (Windows internal process)
7	32560	             "C:\Users\USER\AppData\Local\Programs\Microsoft VS Code\Code.exe" "c:\Users\USER\AppData\Local\Programs\Microsoft VS Code\resources\app\extensions\json-language-features\server\dist\jsonServerMain" --node-ipc --clientProcessId=6624
14	44524	             python refactor.py c:\Users\USER\workspace\project
9	22740	                 console-window-host (Windows internal process)
3	7828	         electron-crash-reporter
6	13076	         watcherService 
2	1716	             console-window-host (Windows internal process)
54	34744	     gpu-process
142	43124	     window (Process Explorer)

[StdioA]

Same on macOS

Every time I use auto completion will create a new Python process, too many of them results in fork failed.

• VS Code version: 1.29.1
• Extension version (available under the Extensions sidebar): 2018.11.0
• OS and version: macOS High Sierra 10.13.6
• Python version (& distribution if applicable, e.g. Anaconda): 3.7.0, homebrew
• Type of virtual environment used (N/A | venv | virtualenv | conda | ...): No