Add identity-audit-check.sh and box identity-audit action

identity-audit-check.sh: proper script replacing the identity-audit-watch
cron inline SSH one-liner. Reads a bl-local cache of the VM audit JSON
(bl cannot SSH to VM; VM hourly audit should push to var/identity-audit.json).
Exits 0 clean, 1 on drift, 2 if cache missing.

box-ctl.py: new identity-audit action returning drift as JSON.
This commit is contained in:
operator-main
2026-10-04 18:35:57 +00:00
parent a76a776a87
commit f679ced960
2 changed files with 90 additions and 0 deletions
+20
View File
@@ -697,6 +697,21 @@ def act_relay_health():
out(all_healthy, relays=results, healthy=all_healthy)
def act_identity_audit():
"""Run identity-audit-check.sh and return drift as JSON."""
audit("identity-audit")
script = BIN / "identity-audit-check.sh"
r = subprocess.run([str(script)], capture_output=True, text=True, timeout=60)
if r.returncode == 2:
fail("AUDIT_UNAVAILABLE", (r.stderr or r.stdout).strip()[:500])
drift = []
for line in r.stdout.strip().split("\n"):
if line.startswith("DRIFT: "):
drift.append(line[len("DRIFT: "):])
clean = r.returncode == 0
out(clean, drift=drift, clean=clean)
def act_cdp_latency():
"""Run cdp-latency-check.sh and return per-node latency JSON."""
@@ -1131,6 +1146,7 @@ fleet:
watchdog-alerts
relay-health
cdp-latency
identity-audit VM identity audit drift check
timer actions:
timer-list
@@ -1363,6 +1379,10 @@ def main(argv):
act_watchdog_alerts()
elif action == "relay-health":
act_relay_health()
elif action == "identity-audit":
if rest:
fail("BAD_ARGS", "usage: identity-audit")
act_identity_audit()
elif action == "cdp-latency":
act_cdp_latency()
elif action == "chrome-errors":
+70
View File
@@ -0,0 +1,70 @@
#!/bin/bash
# identity-audit-check.sh - Check VM identity audit for drift
#
# Lives on bl (/home/super/Projects/NetVM/bin/). Reads a bl-local cached
# copy of the VM's identity audit JSON and reports drift.
#
# Why a cache: bl cannot SSH to the VM (VM only accepts the operator
# container's key). The VM's hourly audit cron (/srv/board/bin/identity-audit.py)
# should scp /srv/board/data/identity-audit.json to bl at:
# /home/super/Projects/NetVM/var/identity-audit.json
# after each run. Until that push is wired, the cache is refreshed manually
# or by the identity-audit-watch cron.
#
# Called by:
# - the identity-audit-watch cron (replaces inline SSH one-liner)
# - box-ctl.py `identity-audit` action
#
# Exit codes:
# 0 - clean (no drift; warnings are expected and silent)
# 1 - drift detected (items printed to stdout, one per line)
# 2 - audit cache missing/unreadable (needs attention)
#
# Output on drift: one drift item per line, prefixed with "DRIFT: "
set -u
CACHE_PATH="/home/super/Projects/NetVM/var/identity-audit.json"
# Allow override for testing
if [ $# -ge 1 ] && [ -f "$1" ]; then
CACHE_PATH="$1"
fi
if [ ! -f "$CACHE_PATH" ]; then
echo "ERROR: identity audit cache missing: $CACHE_PATH" >&2
echo "The VM hourly audit should push /srv/board/data/identity-audit.json here after each run." >&2
exit 2
fi
audit_json=$(cat "$CACHE_PATH" 2>/dev/null)
if [ -z "$audit_json" ]; then
echo "ERROR: identity audit cache unreadable: $CACHE_PATH" >&2
exit 2
fi
# Parse with python3
drift_output=$(printf '%s' "$audit_json" | python3 -c "
import json, sys
try:
data = json.load(sys.stdin)
except Exception as e:
print('ERROR: invalid JSON in audit cache: %s' % e, file=sys.stderr)
sys.exit(2)
for item in data.get('drift', []):
print('DRIFT: ' + str(item))
" 2>&1)
parse_rc=$?
if [ $parse_rc -eq 2 ]; then
echo "$drift_output" >&2
exit 2
fi
if [ -n "$drift_output" ]; then
echo "$drift_output"
exit 1
fi
# Clean: no drift. Warnings are expected (agents not dialed in) — stay silent.
exit 0