feat(creds): Phase 6 — MongoDB SCRAM credential capture
Plugs the cred-coverage gap for MongoDB. The template previously
parsed only the wire opcode + length and discarded the BSON body
entirely, so SCRAM-SHA-{1,256} client-proofs flowed straight through
without ever landing in the Credential table.
Adds an inline minimal BSON walker (~100 LoC) covering the 7 type
codes auth commands actually use: string, doc, array, binary, bool,
int32, int64. Hand-rolled rather than pulling pymongo as a runtime
dep — the parser is bounds-checked for untrusted-input safety
(won't loop on malformed length fields).
Wire flow MongoDB clients use for auth:
- OP_MSG body section (kind=0) → BSON doc with `saslStart` field
carrying mechanism + payload (SCRAM client-first-message:
"n,,n=<user>,r=<nonce>"). Username extracted, pinned to the
per-connection _sasl_username + _sasl_mechanism state.
- Subsequent OP_MSG with `saslContinue` → SCRAM client-final-message
("c=biws,r=<combined>,p=<base64 client-proof>"). The `p=` value is
the credential — emitted as secret_kind=scram_sha256 (or _sha1 /
_unknown depending on the prior saslStart's mechanism), principal
= the pinned username, secret_b64 = base64 of the decoded proof.
Reuse semantics: same client-proof across two auth attempts only
matches when both server salt and password were identical (proofs
include the salt). So cross-session reuse correlates only on
credential reuse against the same MongoDB account on the same decky
— honest, non-misleading signal.
680 tests pass across services, service_testing, db, web/ingester,
and core/fingerprinting (the broader scope my recent commits
touched). Phases 4, 5, 7 still pending (RDP basic-auth, SMB
NTLMSSP, RDP NLA).
This commit is contained in:
@@ -7,12 +7,121 @@ received messages as JSON.
|
||||
"""
|
||||
|
||||
import asyncio
|
||||
import base64
|
||||
import os
|
||||
import struct
|
||||
|
||||
import instance_seed as _seed
|
||||
from syslog_bridge import syslog_line, write_syslog_file, forward_syslog
|
||||
|
||||
|
||||
# ─── Minimal BSON walker ──────────────────────────────────────────────────────
|
||||
# Just enough to extract `saslStart` / `saslContinue` command auth fields.
|
||||
# Pulls a few BSON type codes; ignores everything else (subdocs returned
|
||||
# as raw bytes the caller can re-parse if needed). Hand-rolled rather
|
||||
# than pulling pymongo as a runtime dep — we only need 8 type codes and
|
||||
# the parser is ~40 LoC.
|
||||
|
||||
_BSON_DOUBLE = 0x01
|
||||
_BSON_STRING = 0x02
|
||||
_BSON_DOC = 0x03
|
||||
_BSON_ARRAY = 0x04
|
||||
_BSON_BINARY = 0x05
|
||||
_BSON_BOOL = 0x08
|
||||
_BSON_INT32 = 0x10
|
||||
_BSON_INT64 = 0x12
|
||||
|
||||
|
||||
def _bson_read(buf: bytes, off: int = 0) -> dict:
|
||||
"""Read a single BSON document at ``buf[off]``. Returns a dict of
|
||||
``{key: value}``. Lossy on unsupported types (silently skipped).
|
||||
Untrusted-input safe: bounds-checked, won't infinite-loop on
|
||||
malformed length fields."""
|
||||
out: dict = {}
|
||||
if off + 4 > len(buf):
|
||||
return out
|
||||
doc_len = struct.unpack_from("<i", buf, off)[0]
|
||||
end = off + doc_len
|
||||
if end > len(buf) or doc_len < 5:
|
||||
return out
|
||||
p = off + 4
|
||||
while p < end - 1: # last byte is the trailing 0x00
|
||||
t = buf[p]
|
||||
p += 1
|
||||
if t == 0:
|
||||
break
|
||||
# Read NUL-terminated cstring key.
|
||||
nul = buf.find(b"\x00", p, end)
|
||||
if nul < 0:
|
||||
break
|
||||
key = buf[p:nul].decode("utf-8", errors="replace")
|
||||
p = nul + 1
|
||||
if t == _BSON_STRING:
|
||||
if p + 4 > end:
|
||||
break
|
||||
slen = struct.unpack_from("<i", buf, p)[0]
|
||||
p += 4
|
||||
if p + slen > end or slen < 1:
|
||||
break
|
||||
out[key] = buf[p:p + slen - 1].decode("utf-8", errors="replace")
|
||||
p += slen
|
||||
elif t == _BSON_BINARY:
|
||||
if p + 5 > end:
|
||||
break
|
||||
blen = struct.unpack_from("<i", buf, p)[0]
|
||||
p += 4
|
||||
_subtype = buf[p]
|
||||
p += 1
|
||||
if p + blen > end or blen < 0:
|
||||
break
|
||||
out[key] = buf[p:p + blen] # raw bytes
|
||||
p += blen
|
||||
elif t == _BSON_INT32:
|
||||
if p + 4 > end:
|
||||
break
|
||||
out[key] = struct.unpack_from("<i", buf, p)[0]
|
||||
p += 4
|
||||
elif t == _BSON_INT64:
|
||||
if p + 8 > end:
|
||||
break
|
||||
out[key] = struct.unpack_from("<q", buf, p)[0]
|
||||
p += 8
|
||||
elif t == _BSON_BOOL:
|
||||
if p + 1 > end:
|
||||
break
|
||||
out[key] = buf[p] != 0
|
||||
p += 1
|
||||
elif t == _BSON_DOUBLE:
|
||||
p += 8
|
||||
elif t in (_BSON_DOC, _BSON_ARRAY):
|
||||
if p + 4 > end:
|
||||
break
|
||||
sub_len = struct.unpack_from("<i", buf, p)[0]
|
||||
if p + sub_len > end:
|
||||
break
|
||||
p += sub_len
|
||||
else:
|
||||
# Unsupported type — abort cleanly so we don't misalign.
|
||||
break
|
||||
return out
|
||||
|
||||
|
||||
def _scram_kv(payload: bytes) -> dict:
|
||||
"""Parse a SCRAM message into key=value pairs. SCRAM separates by
|
||||
commas and uses `name=value` pairs. We strip a leading `n,,` (GS2
|
||||
header) when present so the `n=username` shows up directly."""
|
||||
s = payload.decode("utf-8", errors="replace")
|
||||
if s.startswith("n,,"):
|
||||
s = s[3:]
|
||||
elif s.startswith("y,,"):
|
||||
s = s[3:]
|
||||
out: dict = {}
|
||||
for part in s.split(","):
|
||||
if "=" in part:
|
||||
k, _, v = part.partition("=")
|
||||
out[k.strip()] = v
|
||||
return out
|
||||
|
||||
NODE_NAME = os.environ.get("NODE_NAME", "mongodb")
|
||||
SERVICE_NAME = "mongodb"
|
||||
LOG_TARGET = os.environ.get("LOG_TARGET", "")
|
||||
@@ -92,6 +201,11 @@ class MongoDBProtocol(asyncio.Protocol):
|
||||
self._transport = None
|
||||
self._peer = None
|
||||
self._buf = b""
|
||||
# Per-connection SCRAM state: pinned at saslStart so the
|
||||
# subsequent saslContinue's client-proof can carry the username
|
||||
# in the emitted credential row.
|
||||
self._sasl_username: str | None = None
|
||||
self._sasl_mechanism: str | None = None
|
||||
|
||||
def connection_made(self, transport):
|
||||
self._transport = transport
|
||||
@@ -119,6 +233,35 @@ class MongoDBProtocol(asyncio.Protocol):
|
||||
opcode = struct.unpack("<I", msg[12:16])[0]
|
||||
_log("message", src=self._peer[0], opcode=opcode, length=len(msg))
|
||||
|
||||
# SCRAM cred capture: parse the OP_MSG body BSON looking for
|
||||
# saslStart / saslContinue. Each fires its own log event:
|
||||
# saslStart pins the username + mechanism; saslContinue emits
|
||||
# the credential row with the client-proof as secret_b64.
|
||||
if opcode == 2013 and len(msg) >= 21:
|
||||
# OP_MSG body: 4 bytes flagBits, then sections. We only
|
||||
# parse kind=0 (Body) sections — kind=1 (DocSeq) is for
|
||||
# bulk ops that don't carry SCRAM auth.
|
||||
p = 20 # 16 hdr + 4 flagBits
|
||||
while p < len(msg):
|
||||
kind = msg[p]
|
||||
p += 1
|
||||
if kind == 0: # Body section
|
||||
if p + 4 > len(msg):
|
||||
break
|
||||
doc_len = struct.unpack_from("<i", msg, p)[0]
|
||||
if p + doc_len > len(msg):
|
||||
break
|
||||
cmd = _bson_read(msg, p)
|
||||
self._handle_command(cmd)
|
||||
p += doc_len
|
||||
elif kind == 1: # DocSeq — skip
|
||||
if p + 4 > len(msg):
|
||||
break
|
||||
seq_len = struct.unpack_from("<i", msg, p)[0]
|
||||
p += seq_len
|
||||
else:
|
||||
break
|
||||
|
||||
# Build a generic isMaster-style OK response with this instance's
|
||||
# version pair. Fresh topologyVersion on every reply (matches real
|
||||
# mongod behavior — clients use this to detect failover).
|
||||
@@ -141,6 +284,58 @@ class MongoDBProtocol(asyncio.Protocol):
|
||||
else:
|
||||
self._transport.write(_op_reply(request_id, reply_doc))
|
||||
|
||||
def _handle_command(self, cmd: dict) -> None:
|
||||
"""Parse a single MongoDB command document for SCRAM auth.
|
||||
|
||||
saslStart — client-first-message in payload. Extract
|
||||
`n=<username>` so the next saslContinue inherits it.
|
||||
saslContinue — client-final-message in payload. Extract
|
||||
`p=<base64 client-proof>` and emit a cred row.
|
||||
"""
|
||||
# mongo's command dispatch keys off the FIRST field of the BSON
|
||||
# document. We just check key presence since dict ordering in
|
||||
# CPython 3.7+ matches insertion order.
|
||||
if "saslStart" in cmd:
|
||||
mechanism = cmd.get("mechanism")
|
||||
payload = cmd.get("payload") or b""
|
||||
if isinstance(mechanism, str):
|
||||
self._sasl_mechanism = mechanism
|
||||
if isinstance(payload, (bytes, bytearray)):
|
||||
kv = _scram_kv(bytes(payload))
|
||||
self._sasl_username = kv.get("n")
|
||||
_log("auth_start", src=self._peer[0],
|
||||
mechanism=mechanism or "?",
|
||||
username=self._sasl_username or "")
|
||||
return
|
||||
|
||||
if "saslContinue" in cmd:
|
||||
payload = cmd.get("payload") or b""
|
||||
if not isinstance(payload, (bytes, bytearray)):
|
||||
return
|
||||
kv = _scram_kv(bytes(payload))
|
||||
proof_b64 = kv.get("p")
|
||||
if not proof_b64:
|
||||
return
|
||||
try:
|
||||
proof_raw = base64.b64decode(proof_b64, validate=True)
|
||||
except (ValueError, base64.binascii.Error):
|
||||
return
|
||||
mech = (self._sasl_mechanism or "").upper()
|
||||
if "SHA-256" in mech or "SHA256" in mech:
|
||||
kind = "scram_sha256"
|
||||
elif "SHA-1" in mech or "SHA1" in mech:
|
||||
kind = "scram_sha1"
|
||||
else:
|
||||
kind = "scram_unknown"
|
||||
_log("auth", src=self._peer[0],
|
||||
username=self._sasl_username or "",
|
||||
principal=self._sasl_username,
|
||||
mechanism=self._sasl_mechanism or "",
|
||||
secret_kind=kind,
|
||||
secret_printable=proof_b64,
|
||||
secret_b64=base64.b64encode(proof_raw).decode("ascii"))
|
||||
return
|
||||
|
||||
def connection_lost(self, exc):
|
||||
_log("disconnect", src=self._peer[0] if self._peer else "?")
|
||||
|
||||
|
||||
210
tests/services/test_mongodb_scram.py
Normal file
210
tests/services/test_mongodb_scram.py
Normal file
@@ -0,0 +1,210 @@
|
||||
"""MongoDB SCRAM credential capture tests.
|
||||
|
||||
Exercises the inline BSON walker + SCRAM extractor by handcrafting
|
||||
saslStart / saslContinue OP_MSG packets and feeding them to the
|
||||
MongoDBProtocol's data_received(). Asserts that the resulting _log
|
||||
calls carry the universal credential SD shape.
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import base64
|
||||
import importlib.util
|
||||
import struct
|
||||
import sys
|
||||
from pathlib import Path
|
||||
from types import ModuleType
|
||||
from unittest.mock import MagicMock
|
||||
|
||||
|
||||
def _load_mongodb():
|
||||
"""Stand up the mongodb template module with stub deps so the test
|
||||
can poke its protocol directly."""
|
||||
fake = ModuleType("syslog_bridge")
|
||||
fake.syslog_line = MagicMock(return_value="")
|
||||
fake.write_syslog_file = MagicMock()
|
||||
fake.forward_syslog = MagicMock()
|
||||
fake.SEVERITY_INFO = 6
|
||||
fake.SEVERITY_WARNING = 4
|
||||
fake.encode_secret = MagicMock(
|
||||
return_value={"secret_printable": "", "secret_b64": ""}
|
||||
)
|
||||
fake.classify_authorization = MagicMock(return_value=None)
|
||||
sys.modules["syslog_bridge"] = fake
|
||||
|
||||
repo_root = Path(__file__).resolve().parents[2]
|
||||
if "instance_seed" not in sys.modules:
|
||||
spec = importlib.util.spec_from_file_location(
|
||||
"instance_seed", repo_root / "decnet" / "templates" / "instance_seed.py"
|
||||
)
|
||||
seed_mod = importlib.util.module_from_spec(spec)
|
||||
spec.loader.exec_module(seed_mod)
|
||||
sys.modules["instance_seed"] = seed_mod
|
||||
|
||||
spec = importlib.util.spec_from_file_location(
|
||||
"_mongodb_under_test",
|
||||
repo_root / "decnet" / "templates" / "mongodb" / "server.py",
|
||||
)
|
||||
mod = importlib.util.module_from_spec(spec)
|
||||
spec.loader.exec_module(mod)
|
||||
return mod
|
||||
|
||||
|
||||
# ── BSON encoding helpers (test-only) ────────────────────────────────────────
|
||||
|
||||
def _bson_str(key: str, val: str) -> bytes:
|
||||
k = key.encode() + b"\x00"
|
||||
v = val.encode() + b"\x00"
|
||||
return b"\x02" + k + struct.pack("<i", len(v)) + v
|
||||
|
||||
|
||||
def _bson_int32(key: str, val: int) -> bytes:
|
||||
return b"\x10" + key.encode() + b"\x00" + struct.pack("<i", val)
|
||||
|
||||
|
||||
def _bson_binary(key: str, val: bytes) -> bytes:
|
||||
return (
|
||||
b"\x05" + key.encode() + b"\x00"
|
||||
+ struct.pack("<i", len(val))
|
||||
+ b"\x00" # subtype 0 = generic
|
||||
+ val
|
||||
)
|
||||
|
||||
|
||||
def _bson_doc(*fields: bytes) -> bytes:
|
||||
body = b"".join(fields) + b"\x00"
|
||||
return struct.pack("<i", len(body) + 4) + body
|
||||
|
||||
|
||||
def _op_msg(request_id: int, doc: bytes) -> bytes:
|
||||
body = b"\x00\x00\x00\x00" + b"\x00" + doc # flags + kind=0 + body doc
|
||||
return struct.pack("<iiii", 16 + len(body), request_id, 0, 2013) + body
|
||||
|
||||
|
||||
# ── Tests ────────────────────────────────────────────────────────────────────
|
||||
|
||||
def test_bson_walker_basic_types():
|
||||
mod = _load_mongodb()
|
||||
doc = _bson_doc(
|
||||
_bson_str("greeting", "hello"),
|
||||
_bson_int32("answer", 42),
|
||||
_bson_binary("blob", b"\x00\x01\xff"),
|
||||
)
|
||||
parsed = mod._bson_read(doc)
|
||||
assert parsed["greeting"] == "hello"
|
||||
assert parsed["answer"] == 42
|
||||
assert parsed["blob"] == b"\x00\x01\xff"
|
||||
|
||||
|
||||
def test_bson_walker_malformed_safe():
|
||||
mod = _load_mongodb()
|
||||
# Garbage bytes — must not raise or loop.
|
||||
assert mod._bson_read(b"\x05\x00\x00\x00\x00") == {} # 5-byte empty doc
|
||||
assert mod._bson_read(b"\x00" * 4) == {} # too short
|
||||
assert mod._bson_read(b"\xff" * 64) == {} # invalid length
|
||||
|
||||
|
||||
def test_scram_kv_strips_gs2_header():
|
||||
mod = _load_mongodb()
|
||||
payload = b"n,,n=alice,r=clientNonce123"
|
||||
parsed = mod._scram_kv(payload)
|
||||
assert parsed["n"] == "alice"
|
||||
assert parsed["r"] == "clientNonce123"
|
||||
|
||||
|
||||
def test_sasl_start_pins_username():
|
||||
"""saslStart sets per-connection username + mechanism state for the
|
||||
subsequent saslContinue to inherit."""
|
||||
mod = _load_mongodb()
|
||||
proto = mod.MongoDBProtocol()
|
||||
proto._transport = MagicMock()
|
||||
proto._peer = ("10.0.0.5", 51234)
|
||||
|
||||
payload = b"n,,n=alice,r=cnonce"
|
||||
cmd = _bson_doc(
|
||||
_bson_int32("saslStart", 1),
|
||||
_bson_str("$db", "admin"),
|
||||
_bson_str("mechanism", "SCRAM-SHA-256"),
|
||||
_bson_binary("payload", payload),
|
||||
)
|
||||
pkt = _op_msg(request_id=1, doc=cmd)
|
||||
proto.data_received(pkt)
|
||||
|
||||
assert proto._sasl_username == "alice"
|
||||
assert proto._sasl_mechanism == "SCRAM-SHA-256"
|
||||
|
||||
|
||||
def _capture_log(mod):
|
||||
"""Replace mod._log with a list-collector; returns (captured, restore)."""
|
||||
captured: list = []
|
||||
orig = mod._log
|
||||
mod._log = lambda et, severity=6, **kw: captured.append((et, kw))
|
||||
return captured, lambda: setattr(mod, "_log", orig)
|
||||
|
||||
|
||||
def test_sasl_continue_emits_cred():
|
||||
"""saslContinue → emits a _log call with secret_kind="scram_sha256"
|
||||
and secret_b64 = b64(decoded_proof). The _sasl_username pinned in
|
||||
the prior saslStart attaches as principal."""
|
||||
mod = _load_mongodb()
|
||||
proto = mod.MongoDBProtocol()
|
||||
proto._transport = MagicMock()
|
||||
proto._peer = ("10.0.0.5", 51234)
|
||||
proto._sasl_username = "alice"
|
||||
proto._sasl_mechanism = "SCRAM-SHA-256"
|
||||
|
||||
proof = b"\xab" * 32
|
||||
proof_b64 = base64.b64encode(proof).decode("ascii")
|
||||
final_payload = f"c=biws,r=combined,p={proof_b64}".encode()
|
||||
cmd = _bson_doc(
|
||||
_bson_int32("saslContinue", 1),
|
||||
_bson_str("$db", "admin"),
|
||||
_bson_int32("conversationId", 1),
|
||||
_bson_binary("payload", final_payload),
|
||||
)
|
||||
pkt = _op_msg(request_id=2, doc=cmd)
|
||||
|
||||
captured, restore = _capture_log(mod)
|
||||
try:
|
||||
proto.data_received(pkt)
|
||||
finally:
|
||||
restore()
|
||||
|
||||
auth_events = [e for e in captured if e[0] == "auth"]
|
||||
assert len(auth_events) == 1
|
||||
fields = auth_events[0][1]
|
||||
assert fields["secret_kind"] == "scram_sha256"
|
||||
assert fields["principal"] == "alice"
|
||||
assert fields["username"] == "alice"
|
||||
assert base64.b64decode(fields["secret_b64"]) == proof
|
||||
|
||||
|
||||
def test_sasl_continue_unknown_mechanism():
|
||||
"""When mechanism doesn't advertise SHA-{1,256}, fall back to
|
||||
scram_unknown so the row still lands."""
|
||||
mod = _load_mongodb()
|
||||
proto = mod.MongoDBProtocol()
|
||||
proto._transport = MagicMock()
|
||||
proto._peer = ("10.0.0.5", 0)
|
||||
proto._sasl_username = "bob"
|
||||
proto._sasl_mechanism = "PLAIN"
|
||||
|
||||
final_payload = (
|
||||
b"c=biws,r=x,p="
|
||||
+ base64.b64encode(b"proof").decode("ascii").encode()
|
||||
)
|
||||
cmd = _bson_doc(
|
||||
_bson_int32("saslContinue", 1),
|
||||
_bson_str("$db", "admin"),
|
||||
_bson_binary("payload", final_payload),
|
||||
)
|
||||
pkt = _op_msg(request_id=3, doc=cmd)
|
||||
|
||||
captured, restore = _capture_log(mod)
|
||||
try:
|
||||
proto.data_received(pkt)
|
||||
finally:
|
||||
restore()
|
||||
|
||||
auth = [e for e in captured if e[0] == "auth"]
|
||||
assert len(auth) == 1
|
||||
assert auth[0][1]["secret_kind"] == "scram_unknown"
|
||||
Reference in New Issue
Block a user