fix(collector): strip port from remote_addr before attacker identity resolution

host:port in remote_addr was creating a distinct Attacker row per TCP
connection instead of per IP. Split on the last ':' in parse_rfc5424;
preserve the port as fields['remote_port'] so repeated source ports are
retained as fingerprint signal in bounty payloads.
This commit is contained in:
2026-05-10 04:06:42 -04:00
parent 6a6f5807aa
commit f11def0af1
3 changed files with 25 additions and 1 deletions

View File

@@ -472,7 +472,14 @@ def parse_rfc5424(line: str) -> Optional[dict[str, Any]]:
attacker_ip = "Unknown"
for fname in _IP_FIELDS:
if fname in fields:
attacker_ip = fields[fname]
raw = fields[fname]
# remote_addr may be "host:port" — split so identity keys on IP only.
host, _, port = raw.rpartition(":")
if host and port.isdigit():
attacker_ip = host.strip("[]") # handle [::1]:port IPv6 form
fields.setdefault("remote_port", port)
else:
attacker_ip = raw
break
# Fallback for plain `logger` callers that don't use SD params (notably

View File

@@ -640,6 +640,7 @@ async def _extract_bounty(
"protocol": _fields.get("proto") or _fields.get("protocol", "h1"),
"method": _fields.get("method"),
"path": _fields.get("path"),
"remote_port": _fields.get("remote_port"),
},
})
@@ -656,6 +657,7 @@ async def _extract_bounty(
"settings": _fields.get("settings"),
"frame_order": _fields.get("frame_order"),
"protocol": "h2" if _evt_type == "http2_settings" else "h3",
"remote_port": _fields.get("remote_port"),
},
})

View File

@@ -156,6 +156,21 @@ class TestParseRfc5424:
assert result["fields"]["src_ip"] == "1.2.3.4"
assert "src" not in result["fields"]
def test_remote_addr_with_port_strips_port(self):
"""remote_addr="1.2.3.4:40838" — attacker_ip must be the host only."""
line = self._make_line('remote_addr="192.168.1.5:40838"')
result = parse_rfc5424(line)
assert result["attacker_ip"] == "192.168.1.5"
assert result["fields"]["remote_port"] == "40838"
def test_remote_addr_plain_ip_no_port(self):
"""remote_addr="1.2.3.4" without port — attacker_ip is the full value,
no remote_port key injected."""
line = self._make_line('remote_addr="192.168.1.5"')
result = parse_rfc5424(line)
assert result["attacker_ip"] == "192.168.1.5"
assert "remote_port" not in result["fields"]
def test_parses_msg(self):
line = self._make_line(msg="hello world")
result = parse_rfc5424(line)