When I was trying to debug a Python script on cluster, I received the following error: “OpenBLAS blas_thread_init: RLIMIT_NPROC”
So I decided to check why there is OpenBLAS, and quick search leads to numpy, which is a package I used in my script. I quit the program and tried again, no luck.
I typed “top” to find out what is the problem. It showed up that there are around 10 python processes are running but I am not using any at all.
So I used the “-c” option to find the path of each process. It turned out that all the processes are from the VS Code server. It appears some VS Code failed, but the process remains active in the background, and I have never realized that they are there.
Then the question is how to kill them. I used to use “ps” to find running process, but “top” and “ps” are different. I found a solution to:
“pkill -u *” “pkill -9 node”