mitmproxy 11 client replay upstream mode fails #7280
Description
Problem Description
mitmproxy 11 client replay upstream mode fails.
Steps to reproduce the behavior:
- mitmproxy 11.0 crashes doing a client reply of a flow file with upstream proxy set. Client replay works without upstream proxy configured. The same flow file works fine on mitmproxy 10 (it shows some errors, but repeating works)
- Install mitmproxy 10.0
pip install mitmproxy==10
# Create a flow file
mitmdump -p 9000 --save-stream-file foo
[14:34:03.491] HTTP(S) proxy listening at *:9000.
[14:34:16.250][127.0.0.1:52840] client connect
[14:34:16.301][127.0.0.1:52840] server connect google.com:443 (142.250.69.206:443)
127.0.0.1:52840: GET https://google.com/ HTTP/2.0
<< HTTP/2.0 301 Moved Permanently 220b
[14:34:16.392][127.0.0.1:52843] client connect
[14:34:16.414][127.0.0.1:52843] server connect www.google.com:443 (142.251.33.68:443)
127.0.0.1:52843: GET https://www.google.com/ HTTP/2.0
.......SNIPPED........
- Attempt client replay with the captured flow file using an upstream http proxy
mitmdump -nkC foo --mode upstream:http://localhost:8080
[14:34:34.961] [replay] mitmproxy has crashed!
[14:34:34.974] [replay] server connect localhost:8080 (127.0.0.1:8080)
[replay]: GET https://google.com/ HTTP/2.0
<< HTTP/2.0 301 Moved Permanently 220b
[14:34:35.081] [replay] server disconnect localhost:8080 (127.0.0.1:8080)
[14:34:35.082] [replay] mitmproxy has crashed!
[14:34:35.084] [replay] server connect localhost:8080 (127.0.0.1:8080)
[replay]: GET https://www.google.com/ HTTP/2.0
<< HTTP/2.0 200 OK 204k
[14:34:35.284] [replay] server disconnect localhost:8080 (127.0.0.1:8080)
[14:34:35.286] [replay] mitmproxy has crashed!
[14:34:35.288] [replay] server connect localhost:8080 (127.0.0.1:8080)
Although it shows mitmproxy has crashed! the reply works. Also not sure why it show that error.
- Attempt the same with mitmproxy 11
pip install mitmproxy==11
mitmdump -nkC foo --mode upstream:http://localhost:8080
[14:34:48.346] [replay] mitmproxy has crashed!
Error logged during startup, exiting...
[replay]: GET https://google.com/ HTTP/2.0
<< connection cancelled
Shows the same error, but this time exit at first error and does not repeat any request to upstream proxy.
- The client replay work without upstream proxy settings.
mitmdump -nkC foo
[14:35:06.039] [replay] server connect google.com:443 (142.250.69.206:443)
[replay]: GET https://google.com/ HTTP/2.0
<< HTTP/2.0 301 Moved Permanently 220b
[14:35:06.088] [replay] server disconnect google.com:443 (142.250.69.206:443)
[14:35:06.097] [replay] server connect www.google.com:443 (142.251.33.68:443)
[replay]: GET https://www.google.com/ HTTP/2.0
<< HTTP/2.0 200 OK 62.8k
[14:35:06.213] [replay] server disconnect www.google.com:443 (142.251.33.68:443)
[14:35:06.223] [replay] server connect www.google.com:443 (142.251.33.68:443)
....SNIPPED....
I used both BurpSuite and ZAP with intercept mode off as upstream proxy.
System Information
Paste the output of "mitmproxy --version" here.
pip freeze | grep mitm
aioquic-mitmproxy==0.9.21.1
mitmproxy==11.0.0
mitmproxy-macos==0.9.3
mitmproxy_rs==0.9.3
urwid-mitmproxy==2.1.2.1