Circuit list scares relay operators by using term 'exit'

When you attach arm to your shiny new relay, and it does a reachability test by making a circuit back to itself, arm labels the third hop in that circuit the 'Exit'. Which makes at least one user freak out because he thought he set his exit policy to reject :.

Perhaps arm should learn whether circuits are Internal or not (via the getinfo details), and call the final hop something other than Exit in that case?

added component::archived/nyx connections owner::atagar priority::low resolution::duplicate severity::normal status::closed type::enhancement labels

Sounds good. I'm populating that by querying circuit-status and the curses panel receives a list with... (circuitID, status, purpose, (fingerprint1, fingerprint2...))

Can I differentiate these circuits by the purpose field? What would you suggest labeling the final hop as?

Trac:
Status: new to needs_information

Yes: those circuits are of type CIRCUIT_PURPOSE_TESTING: https://gitweb.torproject.org/tor.git/blob/HEAD:/src/or/or.h#l520

You may also find circuit_purpose_to_controller_string() interesting: https://gitweb.torproject.org/tor.git/blob/HEAD:/src/or/circuitlist.c#l336

More generally, you should look at the BUILD_FLAGS part of the circuit-status, for ONEHOP_TUNNEL and IS_INTERNAL. ONEHOP_TUNNEL might be called "dir fetch"; you might call the last hop of an IS_INTERNAL circuit the same thing you call the middle hop (depending I guess on what it is); and you might call the last hop on a testing circuit "reachability test" or "test".

Swapping back to assigned. I'm mostly focused on stem right now but I'm planning to make a sweep through the arm tickets while porting arm over to stem.

Trac:
Status: needs_information to assigned

This is the same ticket as #12956 (closed). Take your pick as to which one is clearer.

Perhaps internal circuits which are not for directory fetches could simply be called "END" hops or "FINAL" hops? Rather than suggesting that anything is actually exiting the network, which seems to be the particular bit of terminology that is frightening folks.

Though I do think that adding multiple new things, one called "DIR FETCH", another called "END", and whatever else is a potentially poor idea in the sense that forcing people to learn too much terminology is overwhelming, and we've already got quite a lot. Perhaps try to pick just one thing, so the reaction doesn't go from "ZOMG, my relay is exiting" to "ZOMG, my relay does I don't even know what". :)

Trac:
Type: defect to enhancement
Priority: Medium to Low
Reviewer: N/A to N/A
Sponsor: N/A to N/A
Severity: N/A to Normal
Keywords: N/A deleted, connections added
Summary: arm circuit list scares relay operators by telling them they're an exit to Circuit list scares relay operators by using term 'exit'

Merging with #12956 (closed).

Trac:
Status: assigned to closed
Resolution: N/A to duplicate

closed

mentioned in issue #12956 (closed)