feat(agent): validate ssh diagnosis receipt
Some checks failed
CD Pipeline / workflow-shape (push) Successful in 0s
CD Pipeline / cancel-stale-cd (push) Has been skipped
CD Pipeline / tests (push) Successful in 36s
CD Pipeline / post-deploy-checks (push) Has been cancelled
CD Pipeline / build-and-deploy (push) Has been cancelled

This commit is contained in:
Your Name
2026-06-30 23:35:49 +08:00
parent e9db7741db
commit bbe051ae2d
5 changed files with 207 additions and 7 deletions

View File

@@ -383,6 +383,16 @@ def _harbor_recovery_receipt_inputs() -> list[dict[str, Any]]:
),
**metadata_boundary,
},
{
"input_id": "ssh_publickey_diagnosis_output",
"source": "diagnose-110-ssh-publickey-auth.sh",
"required_when": "diagnose_ssh_publickey_phase",
"purpose": (
"classify TCP banner, node load, and publickey auth timeout "
"without reading authorized_keys or asking for passwords"
),
**metadata_boundary,
},
{
"input_id": "ssh_local_repair_output",
"source": "recover-110-control-path-and-harbor-local.sh --check/apply",

View File

@@ -21,10 +21,14 @@ def validate_harbor_registry_controlled_recovery_receipt(
if not isinstance(receipt_payload, dict):
raise ValueError("receipt payload must be a JSON object")
ssh_diagnosis_output = _text(
receipt_payload.get("ssh_publickey_diagnosis_output")
)
ssh_local_output = _text(receipt_payload.get("ssh_local_repair_output"))
watchdog_check_output = _text(receipt_payload.get("watchdog_check_output"))
watchdog_repair_output = _text(receipt_payload.get("watchdog_repair_output"))
ssh_diagnosis = _parse_ssh_publickey_diagnosis_output(ssh_diagnosis_output)
ssh_local = _parse_ssh_local_repair_output(ssh_local_output)
watchdog_check = _parse_watchdog_output(watchdog_check_output)
watchdog_repair = _parse_watchdog_output(watchdog_repair_output)
@@ -37,6 +41,7 @@ def validate_harbor_registry_controlled_recovery_receipt(
)
active_blockers = _active_blockers(
ssh_diagnosis=ssh_diagnosis,
ssh_local=ssh_local,
watchdog_check=watchdog_check,
watchdog_repair=watchdog_repair,
@@ -45,6 +50,7 @@ def validate_harbor_registry_controlled_recovery_receipt(
deploy_marker=deploy_marker,
)
status = _status(
ssh_diagnosis=ssh_diagnosis,
ssh_local=ssh_local,
watchdog_check=watchdog_check,
watchdog_repair=watchdog_repair,
@@ -54,6 +60,7 @@ def validate_harbor_registry_controlled_recovery_receipt(
active_blockers=active_blockers,
)
local_console_phase_readback = _local_console_phase_readback(
ssh_diagnosis=ssh_diagnosis,
ssh_local=ssh_local,
watchdog_check=watchdog_check,
watchdog_repair=watchdog_repair,
@@ -71,6 +78,9 @@ def validate_harbor_registry_controlled_recovery_receipt(
"active_blocker_count": len(active_blockers),
"input_redaction": {
"raw_output_returned": False,
"ssh_publickey_diagnosis_output": _text_stats(
ssh_diagnosis_output
),
"ssh_local_repair_output": _text_stats(ssh_local_output),
"watchdog_check_output": _text_stats(watchdog_check_output),
"watchdog_repair_output": _text_stats(watchdog_repair_output),
@@ -87,6 +97,7 @@ def validate_harbor_registry_controlled_recovery_receipt(
},
},
"readback": {
"ssh_publickey_diagnosis": ssh_diagnosis,
"ssh_local_repair": ssh_local,
"watchdog_check": watchdog_check,
"watchdog_repair": watchdog_repair,
@@ -102,6 +113,7 @@ def validate_harbor_registry_controlled_recovery_receipt(
"current_apply_allowed": False,
"current_apply_blocker": _current_apply_blocker(status=status),
"allowed_scope": [
"validate_non_secret_ssh_publickey_diagnosis_receipt",
"validate_non_secret_ssh_local_repair_receipt",
"validate_harbor_watchdog_check_receipt",
"validate_harbor_watchdog_repair_once_receipt",
@@ -123,6 +135,18 @@ def validate_harbor_registry_controlled_recovery_receipt(
},
"learning_writeback_contracts": _learning_writeback_contracts(status=status),
"rollups": {
"ssh_publickey_diagnosis_receipt_seen": ssh_diagnosis[
"receipt_seen"
],
"ssh_publickey_diagnosis_ready": ssh_diagnosis["diagnosis_ready"],
"ssh_publickey_wooo_publickey_classification": ssh_diagnosis[
"wooo_publickey_classification"
],
"ssh_publickey_offer_timeout_seen": ssh_diagnosis[
"publickey_offer_timeout_seen"
],
"ssh_publickey_node_exporter_ok": ssh_diagnosis["node_exporter_ok"],
"ssh_publickey_port_tcp_open": ssh_diagnosis["ssh_port_tcp_open"],
"ssh_local_repair_receipt_seen": ssh_local["receipt_seen"],
"ssh_local_repair_control_channel_metadata_ready": ssh_local[
"control_channel_metadata_ready"
@@ -193,6 +217,7 @@ def validate_harbor_registry_controlled_recovery_receipt(
def _local_console_phase_readback(
*,
ssh_diagnosis: dict[str, Any],
ssh_local: dict[str, Any],
watchdog_check: dict[str, Any],
watchdog_repair: dict[str, Any],
@@ -202,10 +227,10 @@ def _local_console_phase_readback(
_phase(
"diagnose_ssh_publickey",
_phase_status(
ready=ssh_local["receipt_seen"],
ready=ssh_diagnosis["receipt_seen"] or ssh_local["receipt_seen"],
blocked_status="blocked_waiting_ssh_publickey_diagnosis_receipt",
),
"ssh_local_repair",
"ssh_publickey_diagnosis",
),
_phase(
"preflight_control_path_and_harbor",
@@ -217,7 +242,11 @@ def _local_console_phase_readback(
),
_phase(
"repair_ssh_metadata_if_check_confirms_metadata_drift",
_ssh_metadata_phase_status(ssh_local=ssh_local, watchdog_check=watchdog_check),
_ssh_metadata_phase_status(
ssh_diagnosis=ssh_diagnosis,
ssh_local=ssh_local,
watchdog_check=watchdog_check,
),
"ssh_local_repair",
),
_phase(
@@ -268,6 +297,7 @@ def _phase_status(*, ready: bool, blocked_status: str) -> str:
def _ssh_metadata_phase_status(
*,
ssh_diagnosis: dict[str, Any],
ssh_local: dict[str, Any],
watchdog_check: dict[str, Any],
) -> str:
@@ -275,6 +305,8 @@ def _ssh_metadata_phase_status(
if ssh_local["control_channel_metadata_ready"]:
return "ready"
return "blocked_ssh_metadata_repair_receipt_not_ready"
if ssh_diagnosis["publickey_offer_timeout_seen"]:
return "blocked_waiting_ssh_metadata_repair_receipt_after_publickey_timeout"
if watchdog_check["receipt_seen"]:
return "skipped_not_required"
return "blocked_waiting_ssh_metadata_or_harbor_preflight_receipt"
@@ -296,6 +328,77 @@ def _harbor_repair_once_phase_status(
return "blocked_waiting_harbor_watchdog_check_receipt"
def _parse_ssh_publickey_diagnosis_output(output: str) -> dict[str, Any]:
fields = _parse_key_values(output)
marker_seen = "AWOOOI_110_SSH_PUBLICKEY_AUTH_DIAGNOSIS" in output
auth_attempts = _ssh_auth_attempts(output)
wooo_publickey = _auth_classification(
auth_attempts,
user="wooo",
mode="publickey",
)
publickey_offer_timeout_seen = any(
item["mode"] == "publickey"
and item["classification"] == "publickey_offer_timeout"
for item in auth_attempts
)
preauth_timeout_count = sum(
1 for item in auth_attempts if item["classification"] == "preauth_timeout"
)
permission_denied_count = sum(
1 for item in auth_attempts if item["classification"] == "permission_denied"
)
ssh_port_tcp_open = str(fields.get("SSH_PORT") or "") == "tcp_open"
node_exporter_ok = str(fields.get("NODE_EXPORTER") or "") == "ok"
return {
"receipt_seen": marker_seen,
"target": str(fields.get("TARGET") or ""),
"node_exporter_ok": node_exporter_ok,
"node_load_classifier": str(fields.get("NODE_LOAD_CLASSIFIER") or ""),
"node_load1_per_cpu": _float_or_none(fields.get("NODE_LOAD1_PER_CPU")),
"ssh_port_tcp_open": ssh_port_tcp_open,
"ssh_banner_seen": "SSH_BANNER=SSH-" in output,
"auth_attempt_count": len(auth_attempts),
"auth_classifications": auth_attempts,
"wooo_publickey_classification": wooo_publickey,
"publickey_offer_timeout_seen": publickey_offer_timeout_seen,
"preauth_timeout_count": preauth_timeout_count,
"permission_denied_count": permission_denied_count,
"diagnosis_ready": bool(marker_seen and ssh_port_tcp_open and auth_attempts),
"metadata_only": True,
"raw_output_returned": False,
}
def _ssh_auth_attempts(output: str) -> list[dict[str, Any]]:
attempts: list[dict[str, Any]] = []
for line in output.splitlines():
if not line.startswith("SSH_AUTH "):
continue
fields = _parse_key_values(line)
attempts.append(
{
"user": str(fields.get("user") or ""),
"mode": str(fields.get("mode") or ""),
"rc": _int_or_none(fields.get("rc")),
"classification": str(fields.get("classification") or ""),
}
)
return attempts
def _auth_classification(
attempts: list[dict[str, Any]],
*,
user: str,
mode: str,
) -> str:
for item in attempts:
if item["user"] == user and item["mode"] == mode:
return str(item["classification"] or "")
return ""
def _parse_ssh_local_repair_output(output: str) -> dict[str, Any]:
fields = _parse_key_values(output)
marker_seen = "AWOOOI_110_SSH_PUBLICKEY_AUTH_LOCAL_REPAIR" in output
@@ -655,6 +758,7 @@ def _deploy_marker_blockers(
def _active_blockers(
*,
ssh_diagnosis: dict[str, Any],
ssh_local: dict[str, Any],
watchdog_check: dict[str, Any],
watchdog_repair: dict[str, Any],
@@ -663,6 +767,13 @@ def _active_blockers(
deploy_marker: dict[str, Any],
) -> list[str]:
blockers: list[str] = []
if ssh_diagnosis["receipt_seen"] and not ssh_diagnosis["diagnosis_ready"]:
blockers.append("ssh_publickey_diagnosis_receipt_not_ready")
if (
ssh_diagnosis["publickey_offer_timeout_seen"]
and not ssh_local["control_channel_metadata_ready"]
):
blockers.append("ssh_publickey_offer_timeout_on_wooo")
if ssh_local["receipt_seen"] and not ssh_local["control_channel_metadata_ready"]:
blockers.append("ssh_local_repair_receipt_metadata_not_ready")
if not watchdog_check["receipt_seen"]:
@@ -686,6 +797,7 @@ def _active_blockers(
def _status(
*,
ssh_diagnosis: dict[str, Any],
ssh_local: dict[str, Any],
watchdog_check: dict[str, Any],
watchdog_repair: dict[str, Any],
@@ -716,6 +828,8 @@ def _status(
return "harbor_watchdog_check_unhealthy_waiting_repair_once_receipt"
if ssh_local["receipt_seen"]:
return "ssh_local_repair_receipt_waiting_harbor_watchdog_check"
if ssh_diagnosis["receipt_seen"]:
return "ssh_publickey_diagnosis_receipt_waiting_110_local_repair_or_harbor_watchdog_check"
return "blocked_waiting_harbor_controlled_recovery_receipt"
@@ -734,6 +848,11 @@ def _safe_next_step(*, status: str) -> str:
return "run_harbor_watchdog_repair_once_then_submit_receipt_with_verifier"
if status == "ssh_local_repair_receipt_waiting_harbor_watchdog_check":
return "rerun_ssh_control_channel_probe_then_harbor_watchdog_check_mode"
if (
status
== "ssh_publickey_diagnosis_receipt_waiting_110_local_repair_or_harbor_watchdog_check"
):
return "run_110_local_ssh_metadata_check_then_harbor_watchdog_check_mode"
return "submit_non_secret_ssh_local_or_harbor_watchdog_receipt"
@@ -748,6 +867,11 @@ def _current_apply_blocker(*, status: str) -> str:
return "repair_once_receipt_required_after_unhealthy_check"
if status == "ssh_local_repair_receipt_waiting_harbor_watchdog_check":
return "harbor_watchdog_check_receipt_required_after_ssh_local_repair"
if (
status
== "ssh_publickey_diagnosis_receipt_waiting_110_local_repair_or_harbor_watchdog_check"
):
return "110_local_ssh_metadata_or_harbor_watchdog_check_receipt_required"
return "registry_v2_verifier_or_controlled_receipt_required"
@@ -816,6 +940,13 @@ def _int_or_none(value: Any) -> int | None:
return None
def _float_or_none(value: Any) -> float | None:
try:
return float(str(value))
except (TypeError, ValueError):
return None
def _dict(value: Any) -> dict[str, Any]:
return value if isinstance(value, dict) else {}