Files
bf-infra/nautobot/gc-snapshot.py
Christoph Emsenhuber 3877c23505 GC-Backup: täglicher Scheduler angelegt (vorher gab es keinen)
Diagnose: die GC-Backups standen seit 2026-09-04, weil NIE ein Scheduler
existierte — Backup/Intended/Compliance liefen nur manuell aus Sessions. Die
"2×/Tag per SSH"-Annahme (CLAUDE.md Bot-Master-Zeile + gc-snapshot.py-Docstring)
war nie real; beide korrigiert.

Fix: Nautobot-Job "Execute All Golden Configuration Jobs - Multiple Device"
enabled + ScheduledJob "GC Backup+Compliance – taeglich" (platform=RouterOS,
täglich 03:30 UTC, Backup→Intended→Compliance). Sofort-Lauf SUCCESS, alle 5
Geräte backup/compliance auf 2026-09-12 10:35, frischer bf-golden-config-Commit.
Announcement-Eintrag ergänzt.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-09-12 12:36:48 +02:00

200 lines
8.1 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env python3
"""
TP36-Live — Live-Gerätedaten als CustomFields auf den RouterOS-dcim.Device.
Legt (idempotent, get-or-create) fünf CustomFields auf `dcim.device` an und
befüllt sie für jedes RouterOS-Gerät. Die Werte kommen AUSSCHLIESSLICH aus
Nautobot selbst — es wird KEIN Gerät kontaktiert:
ros_version aus der GoldenConfig-`backup_config` ("by RouterOS X.Y.Z")
dhcp_server_count gezählte Einträge der Sektion "/ip dhcp-server" im Backup
dns_upstream_mode aus der "/ip dns"-Sektion des Backups: doh | plain | none
compliance_pct GC config-compliance: compliant / total * 100 (gerundet)
mgmt_ip device.primary_ip4 (ohne /prefix)
Warum aus Nautobot statt live vom Gerät: der MGMT-/REST-Weg (VLAN99:80) ist
nicht aus jedem Vantage erreichbar; die GC-Backups werden täglich (03:30 UTC,
ScheduledJob "GC Backup+Compliance – taeglich") per SSH gezogen und sind die
Doku-SSoT. Damit braucht der Snapshot keine Geräte-Credentials. Aktualität steht
in `backup_last_success_date` — ist die alt, steht der Backup-Schedule (2026-09-12
war er 8 Tage aus, weil es vorher gar keinen gab).
Der Backup-Config-Text wird nur INTERN geparst (Zähler/Modus abgeleitet) und
NIE ausgegeben — er kann gescrubte Reste enthalten (PSK-Leak-Regel).
Token-Quelle: docs/nas-vlan20/stacks/210-nautobot.yml (gitignored),
Schlüssel NAUTOBOT_SUPERUSER_API_TOKEN.
Aufruf:
python3 nautobot/gc-snapshot.py # Dry-Run (default) — zeigt nur an
python3 nautobot/gc-snapshot.py --commit # legt CustomFields an + patcht Devices
"""
import json
import os
import re
import sys
import urllib.request
import urllib.parse
import urllib.error
from pathlib import Path
NB = os.environ.get("NAUTOBOT_URL", "http://10.40.20.30:8080")
REPO = Path(__file__).resolve().parent.parent
YML = REPO / "docs" / "nas-vlan20" / "stacks" / "210-nautobot.yml"
COMMIT = "--commit" in sys.argv
def load_token() -> str:
for line in YML.read_text().splitlines():
m = re.search(r"NAUTOBOT_SUPERUSER_API_TOKEN[=:]\s*([^\s\"']+)", line)
if m:
return m.group(1)
sys.exit(f"NAUTOBOT_SUPERUSER_API_TOKEN nicht in {YML} gefunden")
TOKEN = load_token()
HDRS = {"Authorization": f"Token {TOKEN}", "Accept": "application/json",
"Content-Type": "application/json"}
def req(method: str, path: str, body=None):
data = json.dumps(body).encode() if body is not None else None
r = urllib.request.Request(f"{NB}{path}", data=data, headers=HDRS, method=method)
try:
with urllib.request.urlopen(r, timeout=20) as resp:
return resp.status, json.loads(resp.read().decode() or "{}")
except urllib.error.HTTPError as e:
return e.code, json.loads(e.read().decode() or "{}")
# --------------------------------------------------------------- CustomFields
# (key, label, type, description). content_types wird ergänzt.
CUSTOM_FIELDS = [
("ros_version", "RouterOS-Version", "text",
"RouterOS-Version aus der GC-backup_config (by RouterOS …)."),
("compliance_pct", "GC-Compliance %", "integer",
"Golden-Config config-compliance: compliant/total in Prozent."),
("dhcp_server_count", "DHCP-Server", "integer",
"Anzahl Einträge der Sektion /ip dhcp-server im GC-Backup."),
("mgmt_ip", "Mgmt-IP", "text",
"device.primary_ip4 ohne Prefix (VLAN99-Management)."),
("dns_upstream_mode", "DNS-Upstream-Modus", "text",
"Aus /ip dns des GC-Backups: doh | plain | none."),
]
CF_KEYS = [c[0] for c in CUSTOM_FIELDS]
def ensure_custom_fields():
"""get-or-create die fünf CFs; bei Bestand: content_type dcim.device sicherstellen."""
st, d = req("GET", "/api/extras/custom-fields/?limit=0")
have = {r["key"]: r for r in d.get("results", [])}
for key, label, typ, desc in CUSTOM_FIELDS:
ex = have.get(key)
if ex:
cts = ex.get("content_types", [])
ok = "dcim.device" in cts
print(f" = CF '{key}' existiert ({ex['id']}){'' if ok else ' ⚠ ohne dcim.device!'}")
if not ok and COMMIT:
req("PATCH", f"/api/extras/custom-fields/{ex['id']}/",
{"content_types": sorted(set(cts) | {"dcim.device"})})
print(f" + dcim.device ergänzt")
continue
payload = {"key": key, "label": label, "type": typ, "description": desc,
"content_types": ["dcim.device"], "grouping": "TP36-Live",
"required": False}
if not COMMIT:
print(f" + CF '{key}' ({typ}) [DRY-RUN]")
continue
st, r = req("POST", "/api/extras/custom-fields/", payload)
print(f" {'+' if st in (200,201) else '!'} CF '{key}'"
+ (f" angelegt ({r.get('id')})" if st in (200, 201)
else f" FEHLER {st}: {json.dumps(r)[:200]}"))
# --------------------------------------------------------------- Backup-Parser
def parse_backup(cfg: str):
"""Werte aus dem flachen RouterOS-Export (eine '/…'-Zeile pro Kommando)."""
ver = ""
m = re.search(r"by RouterOS ([0-9][0-9A-Za-z.\-]*)", cfg)
if m:
ver = m.group(1)
# DHCP-Server = Zeilen '/ip dhcp-server add …' (NICHT lease/network/option)
dhcp = len(re.findall(r"(?m)^\s*/ip dhcp-server add\b", cfg))
# DNS-Modus aus der '/ip dns set …'-Zeile
mode = "none"
md = re.search(r"(?m)^\s*/ip dns set\b.*", cfg)
if md:
line = md.group(0)
doh = re.search(r"use-doh-server=([^\s]+)", line)
srv = re.search(r"\bservers=([^\s]+)", line)
if doh and doh.group(1) not in ("", '""'):
mode = "doh"
elif srv and srv.group(1) not in ("", '""'):
mode = "plain"
return ver, dhcp, mode
def compliance_pct(device_name: str):
q = urllib.parse.urlencode({"device": device_name, "limit": 0, "depth": 0})
st, d = req("GET", f"/api/plugins/golden-config/config-compliance/?{q}")
rows = d.get("results", [])
if not rows:
return None
ok = sum(1 for r in rows if r.get("compliance"))
return round(ok / len(rows) * 100)
# ---------------------------------------------------------------------- main
def routeros_devices():
q = urllib.parse.urlencode({"limit": 0, "depth": 1, "platform": "RouterOS"})
st, d = req("GET", f"/api/dcim/devices/?{q}")
devs = []
for r in d.get("results", []):
ip = r.get("primary_ip4")
ip = ip.get("address") if isinstance(ip, dict) else ip
if not ip:
continue # tote Altgeräte (cudy/sw) ohne IP überspringen
devs.append((r["id"], r["name"], ip.split("/")[0]))
return sorted(devs, key=lambda x: x[1])
def main():
print(f"== TP36-Live gc-snapshot ({'COMMIT' if COMMIT else 'DRY-RUN'}) @ {NB} ==")
print("\n[1] CustomFields sicherstellen")
ensure_custom_fields()
print("\n[2] Golden-Config-Backups laden")
st, gc = req("GET", "/api/plugins/golden-config/golden-config/?depth=1&limit=0")
backups = {}
for r in gc.get("results", []):
dev = r.get("device")
dev = dev.get("name") if isinstance(dev, dict) else dev
backups[dev] = (r.get("backup_config") or "", r.get("backup_last_success_date") or "?")
print("\n[3] Snapshot je Gerät")
print(f" {'Gerät':16} {'ros':9} {'cmpl%':>5} {'dhcp':>4} {'dns':6} {'mgmt-ip':15} backup")
for dev_id, name, mgmt in routeros_devices():
cfg, when = backups.get(name, ("", "—"))
ver, dhcp, dns = parse_backup(cfg) if cfg else ("?", 0, "?")
pct = compliance_pct(name)
vals = {
"ros_version": ver or None,
"compliance_pct": pct,
"dhcp_server_count": dhcp,
"mgmt_ip": mgmt,
"dns_upstream_mode": dns,
}
pct_s = "—" if pct is None else str(pct)
print(f" {name:16} {ver or '?':9} {pct_s:>5} {dhcp:>4} {dns:6} {mgmt:15} {when[:10]}")
if COMMIT:
st, r = req("PATCH", f"/api/dcim/devices/{dev_id}/", {"custom_fields": vals})
if st not in (200, 201):
print(f" ! PATCH FEHLER {st}: {json.dumps(r)[:200]}")
print(f"\n== fertig ({'COMMIT — geschrieben' if COMMIT else 'DRY-RUN — nichts geschrieben'}) ==")
if __name__ == "__main__":
main()