Files
securityonion/salt/_beacons/postgres_pillar_beacon.py
Josh Patterson 53f9ebcd46 FIX: queue auto-applied state runs instead of failing on conflict
orch.push_batch passed `kwarg: {queue: 2}` to salt.state, but in Salt
3006 queue is a top-level salt.state argument and salt.state always sets
the minion's queue kwarg from it (default False), so the kwarg block was
silently dropped and every pushed state ran with queue=False. The drainer
dispatches a separate async orchestration each 15s pass, so settings saved
more than ~15s apart overlap on the same minion and every run after the
first fails immediately with 'The function "state.sls" is running as PID
...'. The change then waits for the next scheduled highstate.

Seen on a 3.4.0 standalone: hydra.enabled, telegraf.output, and two soc
settings (including soc.config.licenseKey) were saved within 30s. The soc
state was dispatched while the telegraf state was still running and was
rejected, so the license key was not applied.

Use `queue: True`, as orch.deploy_newnode already does. An int is treated
as max_queue and still falls through to the conflict error once that many
state runs are active.

The failure was only visible in the master log, since the drainer
dispatches with --async and logged only "dispatch accepted". The drainer
now:
  - logs each dispatched action
  - parses the orchestration jid from salt-run's stderr (the only place
    --async reports it) and records it under /opt/so/state/push_dispatched
  - on later passes looks each jid up with jobs.lookup_jid and logs either
    "push succeeded" or an ERROR with the failed step, the per-minion
    failed states or rejection text, and the triggering paths
Lookups run outside the pending-intent lock since the reactors share it.

The beacon now logs each audit_settings row it emits and the reactor logs
the audit row id, so a single change can be traced from audit_settings to
its push result.

Adds so-push-drainer_test.py; the drainer is now held to the 100% coverage
requirement in python-test.

Verified on the standalone: a soc push dispatched while a 90s state run
was in progress queued behind it (queue=True in the job args), completed,
and the drainer logged "push succeeded" for its jid. The new result
parsing reports the original soc conflict and the hydra license failure
from the job cache.
2026-09-30 16:18:37 -04:00

145 lines
5.2 KiB
Python

# Copyright Security Onion Solutions LLC and/or licensed to Security Onion Solutions LLC under one
# or more contributor license agreements. Licensed under the Elastic License 2.0 as shown at
# https://securityonion.net/license; you may not use this file except in compliance with the
# Elastic License 2.0.
# Custom salt beacon that watches the SOC audit_settings table in postgres for
# new settings changes and emits a beacon event per new row. This replaces the
# inotify watch on /opt/so/saltstack/local/pillar -- instead of monitoring pillar
# files on disk, we monitor the securityonion.audit_settings table that SOC writes to.
#
# Detection is poll-based with a monotonic `id` watermark persisted to
# WATERMARK_FILE: each pass selects rows with id greater than the last id seen,
# which makes it self-healing (a missed poll simply catches up on the next one).
#
# Each emitted event carries setting_id and node_id; the push_pillar reactor maps
# setting_id -> app via pillar_push_map.yaml and writes a push intent, after which
# the existing so-push-drainer / orch.push_batch pipeline takes over unchanged.
import logging
import os
import subprocess
log = logging.getLogger(__name__)
WATERMARK_FILE = '/opt/so/state/postgres_pillar_beacon_watch.id'
CONTAINER = 'so-postgres'
DATABASE = 'securityonion'
# Unaligned, tuples-only psql output with a field separator that cannot appear in
# an id/setting_id/node_id, so we can split each row reliably.
FIELD_SEP = '\x1f'
def __virtual__():
return True
def validate(config):
return True, 'valid'
def _read_watermark():
# Returns the last processed id, or None if the watermark has not been seeded.
try:
with open(WATERMARK_FILE, 'r') as f:
return int((f.read() or '').strip())
except (IOError, ValueError):
return None
def _write_watermark(value):
try:
os.makedirs(os.path.dirname(WATERMARK_FILE), exist_ok=True)
tmp = WATERMARK_FILE + '.tmp'
with open(tmp, 'w') as f:
f.write(str(int(value)))
os.rename(tmp, WATERMARK_FILE)
except OSError:
log.exception('postgres_pillar_beacon: failed to persist watermark to %s', WATERMARK_FILE)
def _query(sql):
# Run a query against securityonion inside the so-postgres container over the unix
# socket (trust auth, no password). Returns stdout on success, or None on any
# failure so the caller can no-op and retry on the next interval.
cmd = [
'docker', 'exec', CONTAINER,
'psql', '-U', 'postgres', '-d', DATABASE,
'-tA', '-F', FIELD_SEP, '-c', sql,
]
try:
result = subprocess.run(cmd, capture_output=True, text=True, timeout=30)
except subprocess.TimeoutExpired:
log.warning('postgres_pillar_beacon: psql timed out')
return None
except Exception:
log.exception('postgres_pillar_beacon: failed to exec psql')
return None
if result.returncode != 0:
log.warning('postgres_pillar_beacon: psql failed (rc=%s): %s',
result.returncode, (result.stderr or '').strip())
return None
return result.stdout
def beacon(config): # noqa: C901
retval = []
watermark = _read_watermark()
# First run / missing watermark: seed to the current MAX(id) and emit nothing
# so we never replay the entire settings history into a fleetwide push.
if watermark is None:
seed = _query('SELECT COALESCE(MAX(id), 0) FROM audit_settings;')
if seed is None:
return retval # postgres not ready yet; retry next interval
try:
_write_watermark(int((seed or '0').strip() or 0))
except ValueError:
log.warning('postgres_pillar_beacon: could not parse MAX(id) seed: %r', seed)
return retval
rows = _query(
"SELECT id, setting_id, COALESCE(node_id, '') FROM audit_settings "
"WHERE id > %d ORDER BY id;" % watermark
)
if rows is None:
return retval
max_id = watermark
for line in rows.splitlines():
# Do NOT str.strip() the whole line: Python treats the \x1f field
# separator (and \x1c-\x1e) as whitespace, so stripping would eat an
# empty trailing node_id field and make the row look malformed.
if not line.strip():
continue
parts = line.split(FIELD_SEP)
if len(parts) < 3:
log.warning('postgres_pillar_beacon: skipping malformed row: %r', line)
continue
try:
row_id = int(parts[0])
except ValueError:
log.warning('postgres_pillar_beacon: skipping row with non-int id: %r', line)
continue
setting_id = parts[1]
node_id = parts[2]
retval.append({
'tag': 'audit_settings',
'id': row_id,
'setting_id': setting_id,
'node_id': node_id,
})
log.info('postgres_pillar_beacon: audit_settings id=%d setting_id=%s node_id=%s',
row_id, setting_id, node_id)
if row_id > max_id:
max_id = row_id
if max_id > watermark:
_write_watermark(max_id)
log.info('postgres_pillar_beacon: emitted %d change(s), watermark %d -> %d',
len(retval), watermark, max_id)
return retval