anti
4c37ece39e
Adds a new decnet orchestrate worker whose job is to keep the honeypot
ecosystem from looking suspiciously static — a frozen LAN with no
inter-host traffic and no filesystem aging is its own honeypot tell.
MVP scope:
- New OrchestratorEvent table + repo methods (purpose-built sibling
to Log so synthetic events stay separable from attacker-driven ones).
- New orchestrator.{activity,file}.<decky_id> bus topics +
system.orchestrator.health heartbeat.
- SSH-only driver. Traffic action runs python3 inside src container
to TCP-connect dst:22 and read the SSH banner — real on-the-wire
SSH-protocol traffic without shipping creds. File action drops or
refreshes a small file via docker exec on the destination.
- Random scheduler (50/50 traffic/file when >=2 SSH-capable deckies
are running). Diurnal shaping, role-aware pairing, and session-aware
backoff are explicit non-goals for MVP.
- CLI registration, systemd unit (SupplementaryGroups=docker),
worker-registry entry so the dashboard shows orchestrator health.
- 11 tests: scheduler policy, driver argv shape + injection-safety,
end-to-end one-tick integration with FakeBus + SQLite.
115 lines
3.7 KiB
Python
115 lines
3.7 KiB
Python
"""Orchestrator main loop.
|
|
|
|
One tick = one (src, dst, action) pick + one driver invocation + one DB
|
|
write + one fire-and-forget bus publish. Intentionally serial — MVP
|
|
honesty: a wedged docker exec stalls only this worker, never another.
|
|
|
|
Modeled after :mod:`decnet.profiler.worker` for consistency: same control
|
|
listener, same heartbeat helper, same shutdown semantics.
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import asyncio
|
|
import contextlib
|
|
|
|
from decnet.bus.factory import get_bus
|
|
from decnet.bus.publish import (
|
|
publish_safely,
|
|
run_control_listener,
|
|
run_health_heartbeat,
|
|
)
|
|
from decnet.logging import get_logger
|
|
from decnet.orchestrator import events, scheduler
|
|
from decnet.orchestrator.drivers import SSHDriver
|
|
from decnet.web.db.repository import BaseRepository
|
|
|
|
logger = get_logger("orchestrator")
|
|
|
|
|
|
async def orchestrator_worker(
|
|
repo: BaseRepository,
|
|
*,
|
|
interval: int = 60,
|
|
) -> None:
|
|
"""Periodically inject synthetic activity into the running fleet.
|
|
|
|
Runs as a long-lived asyncio task. Honours the bus control topic
|
|
(``system.orchestrator.control``) for graceful shutdown.
|
|
"""
|
|
logger.info("orchestrator worker started interval=%ds", interval)
|
|
|
|
bus = None
|
|
try:
|
|
bus = get_bus(client_name="orchestrator")
|
|
await bus.connect()
|
|
except Exception as exc: # noqa: BLE001
|
|
logger.warning(
|
|
"orchestrator: bus unavailable, continuing without publish: %s", exc
|
|
)
|
|
bus = None
|
|
|
|
driver = SSHDriver()
|
|
shutdown = asyncio.Event()
|
|
heartbeat_task = asyncio.create_task(run_health_heartbeat(bus, "orchestrator"))
|
|
control_task = asyncio.create_task(
|
|
run_control_listener(bus, "orchestrator", shutdown),
|
|
)
|
|
try:
|
|
while not shutdown.is_set():
|
|
try:
|
|
await asyncio.wait_for(shutdown.wait(), timeout=interval)
|
|
except asyncio.TimeoutError:
|
|
pass # normal tick
|
|
if shutdown.is_set():
|
|
break
|
|
try:
|
|
await _one_tick(repo, driver, bus)
|
|
except Exception as exc: # noqa: BLE001
|
|
logger.error("orchestrator tick failed: %s", exc)
|
|
finally:
|
|
for t in (heartbeat_task, control_task):
|
|
t.cancel()
|
|
with contextlib.suppress(Exception, asyncio.CancelledError):
|
|
await t
|
|
if bus is not None:
|
|
with contextlib.suppress(Exception):
|
|
await bus.close()
|
|
|
|
|
|
async def _one_tick(repo: BaseRepository, driver, bus) -> None:
|
|
deckies = await repo.list_running_topology_deckies()
|
|
action = scheduler.pick(deckies)
|
|
if action is None:
|
|
logger.debug(
|
|
"orchestrator: no actionable deckies (running+ssh count=%d)",
|
|
len(deckies),
|
|
)
|
|
return
|
|
|
|
result = await driver.run(action)
|
|
row = events.to_row(action, result)
|
|
await repo.record_orchestrator_event(row)
|
|
|
|
if bus is not None:
|
|
topic = events.topic_for(action)
|
|
# Bus payload mirrors the row but uses iso8601 for ts so SSE
|
|
# consumers don't have to JSON-handle datetime themselves.
|
|
bus_payload = {
|
|
"kind": row["kind"],
|
|
"protocol": row["protocol"],
|
|
"action": row["action"],
|
|
"src_decky_uuid": row.get("src_decky_uuid"),
|
|
"dst_decky_uuid": row["dst_decky_uuid"],
|
|
"success": row["success"],
|
|
"payload": result.payload,
|
|
"ts": row["ts"].isoformat(),
|
|
}
|
|
await publish_safely(
|
|
bus, topic, bus_payload, event_type=events.event_type_for(action)
|
|
)
|
|
|
|
logger.info(
|
|
"orchestrator tick kind=%s success=%s dst=%s",
|
|
row["kind"], row["success"], row["dst_decky_uuid"],
|
|
)
|