"Address already in use" even when it's not

Hi

I am running multiple mitmproxy instances on one machine. Each instance has its own port between 49000 and 50000. All instances are listening on interface 0.0.0.0. Instances stop and restart multiple times per hour.

Most instances start fine. However, a few random ones will fail with the error that address is already in use:

$ mitmproxy -p 49338 Error starting proxy server: error(98, 'Address already in use')

However, netstat does not show any process using the port:

$ sudo netstat -nlp | grep 49338

I also do not see any zombie processes trying to start the same instance.

Any idea what’s causing this? Is the problem that every instance is listening on 0.0.0.0? How does mitmproxy decide whether a port is “open”?

[Kaushal0812]

Just in case above solutions didn’t work:

Follows the below steps:

1. Get the port your process is listening to: $ ps ax | grep python

2. Kill the Process $ kill PROCESS_NAME

[hynguyen2610]

I ran into this problem too an not found any solution yet. Someone said about ps -f ax | grep mitm I tried an kill the processes in ps command result but mitmproxy still can not be started. The error message is the same.