Within the last night, Rasa used 100% disk space of my server. It created many files in /tmp, around ~80GB.
I think it has something to do with the error, if the address is already in use:
Everytime Rasa tries to start, a small amount of disk space is used. It’s about ~200MB/minute.It’s started with this command:
rasa run --endpoints endpoints.yml --enable-api --credentials credentials.yml --cors "*" -p 5050
I had the action server on the same port, but changed it now. I just wanted to report the bug, that it fills the disk with temporary files, if the port is occupied.
is there any solution to this, i am also facing alot of memory issue
I was also facing the same issue. rasa was generating big size files every minute.
What happened in the background is:
Whenever you first run the rasa with the command: rasa run --endpoints endpoints.yml --enable-api --credentials credentials.yml --cors "*" -p 5050
First, rasa generates and stores the files in the /tmp folder which is required later to perform the operations.
And then look for a port which is specified: 5050 in this case or by default 5005, if this port is free then the rasa server will run on the same port.
and if your rasa server is already up and running and even if you try to run the rasa on the same port: in this case, it is 5050,
then rasa first stores the files inside the /tmp folder and checks for the port whether it is free or not?
This is the main reason why rasa generates multiple files
imp: On the deployed server, if your script is written in such a way that it will run the above command multiple times for the same port until the command runs successfully, then it will generate multiple files.
one more thing: whenever you stop the rasa server and free the respective port, it will automatically vanish the data from the /tmp folder.
correct me if I am wrong. Thanks!
