"SSH is using the default SocksPort" test suite scenario is fragile
I’ve initially spotted this on
Bug #16792, and then noticed that it affected all branches.
For example, this scenario failed 12 times in the last 16 test suite runs on the testing branch.
In every failed case, SSH immediately fails with “nc: connection failed, SOCKSv5 error: General SOCKS server failure”.
Interestingly, none of the scenarios in
ssh.feature are affected, while they do almost the same thing. The main differences are:
tor_stream_isolation.featurewe have the “I monitor the network connections of SSH” CPU hog running. This might make it harder for tor to do its job.
ssh.feature, we use
retry_torso we retry up to
MAX_NEW_TOR_CIRCUIT_RETRIES(default: 10); while in
tor_stream_isolation.featurewe only try once.
In short, the fragile scenario runs in a context that makes it more likely to fail, and does not retry.