There seems to be an issue with duplicate key detection

[claudijd]

In scanning a large number of hosts I'm seeing duplicate keys where duplicate keys don't actually exist.

Need to dig in and figure out why this is happening.

This can be seen here:

https://github.com/mozilla/ssh_scan/blob/master/examples/github.ghproxy.top.json#L90-L93

[claudijd]

This is due to the addition of the "known_keys" attribute being included in what was previously thought to only include key values, so the hostkey db is polluted.

[claudijd]

Problem code exists here:

https://github.com/mozilla/ssh_scan/blob/master/lib/ssh_scan/scan_engine.rb#L152-L156

And here:

https://github.com/mozilla/ssh_scan/blob/master/lib/ssh_scan/scan_engine.rb#L165-L171

[claudijd]

The good thing here is that polluted records should self-correct in a couple ways based on #352

1.) When scanning IP A it will remove the old fingerprint and apply the new non-polluted one.
2.) As a result of scanning IP A it will no longer detect an arbitrary collision with IP B