Cam Proudlock
870e068c44
The board was one card per source in a fixed grid: every source got the same small box whatever it had to say (warranties showed 5 of 50), and a box per source flattened priority, while the largest type on the page belonged to counts that need nobody. Now a severity rollup, one worst-first queue of everything needing a person, and a row of context beneath. The queue is full width, capped per source so a noisy check cannot bury a rare one, and names what did not fit with a link to it. Chips filter the queue by check as well as by severity; both narrow what is already loaded, so the counts and rows cannot disagree. Dropped: an 'All assets' tile shown beside its own four addends, a printer strip carrying the same count a third time, three cards whose job was to display 0, and a Recent Devices table listing five machines all 'In Use'. Contract 0.26.0 adds optional 'shortname' on dashboard cards - a title is written as a heading and ran wider than the hostname it qualified as a row badge. Also: the supplies card reported 'Low Supplies 0' when Zabbix was unreachable, which reads as reassurance rather than the truth that nothing could see the cartridges. It now distinguishes unreachable (says so) from unconfigured (stays quiet - the site declined the feature) from a real zero. Full suite 2182 passed, frontend 314 passed.
522 lines
24 KiB
Python
522 lines
24 KiB
Python
"""Backups plugin main class.
|
|
|
|
Per-asset configuration backups with revision history, pluggable by kind
|
|
(services/registry.py). Accepts collector input over HTTPS per ADR-006 and
|
|
renders one asset panel per kind via the ADR-010 hook.
|
|
|
|
Dedup is the load-bearing behaviour: GE-Enforce runs the collector every cycle
|
|
across the fleet, so apply_collector_payload inserts a revision only when the
|
|
content hash differs from that asset's latest for that kind. Without it a
|
|
readable history would be buried under millions of identical rows.
|
|
"""
|
|
|
|
import base64
|
|
import json
|
|
import logging
|
|
from datetime import datetime, timezone
|
|
from pathlib import Path
|
|
from typing import Dict, List, Optional, Type
|
|
|
|
from flask import Flask, Blueprint
|
|
|
|
from shopdb.plugins.base import BasePlugin, PluginMeta
|
|
from shopdb.api import Setting
|
|
|
|
from .api import backups_bp
|
|
from .models import BackupRevision
|
|
from .services.registry import (REGISTRY, getkind, canonicalhash,
|
|
DEFAULTSHAREROOT)
|
|
from .services.retention import prune, samesource as _samesource
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
class BackupsPlugin(BasePlugin):
|
|
"""Configuration backup + revision history plugin."""
|
|
|
|
def __init__(self):
|
|
self._manifest = self._load_manifest()
|
|
|
|
def _load_manifest(self) -> Dict:
|
|
manifest_path = Path(__file__).parent / 'manifest.json'
|
|
if manifest_path.exists():
|
|
with open(manifest_path, 'r') as f:
|
|
return json.load(f)
|
|
return {}
|
|
|
|
@property
|
|
def meta(self) -> PluginMeta:
|
|
return PluginMeta(
|
|
name=self._manifest.get('name', 'backups'),
|
|
version=self._manifest.get('version', '1.0.0'),
|
|
description=self._manifest.get('description', 'Configuration backups'),
|
|
author=self._manifest.get('author', 'ShopDB Team'),
|
|
dependencies=self._manifest.get('dependencies', []),
|
|
core_version=self._manifest.get('core_version', '>=0.2.0,<1.0.0'),
|
|
api_prefix=self._manifest.get('api_prefix', '/api/backups'),
|
|
)
|
|
|
|
def get_blueprint(self) -> Optional[Blueprint]:
|
|
return backups_bp
|
|
|
|
def get_models(self) -> List[Type]:
|
|
return [BackupRevision]
|
|
|
|
def get_permissions(self) -> List:
|
|
"""RBAC for the human-facing routes.
|
|
|
|
Deliberately split: download is separate from view because a .reg is a
|
|
restorable artifact, so a read-only auditor should be able to see that a
|
|
machine's config changed without being able to pull a file that
|
|
reconfigures it.
|
|
|
|
No delete permission. Revisions are removed only by the retention
|
|
policy; a hand-delete route would let someone quietly drop the evidence
|
|
of a config change, and there is no operational need for it. Collector
|
|
INGEST is not covered here at all - that path authorizes on a
|
|
collector.ingest-scoped service token, not a user JWT.
|
|
"""
|
|
return [
|
|
('backups.view', 'View configuration backups', 'backups'),
|
|
('backups.download', 'Download configuration backups', 'backups'),
|
|
]
|
|
|
|
def get_asset_panels(self) -> List[Dict]:
|
|
"""One panel per kind, plus the DNC Info card.
|
|
|
|
DNC Info sits ABOVE the history panels (lower position) because it
|
|
answers the question a tech actually arrives with - what is this
|
|
machine's controller talking to - while the history panels are for the
|
|
rarer restore case.
|
|
"""
|
|
panels = [kind.infopanel() for kind in REGISTRY.values()
|
|
if kind.infopanel()]
|
|
for position, kind in enumerate(REGISTRY.values()):
|
|
panel = {
|
|
'id': 'backups-{}'.format(kind.key),
|
|
'title': kind.displayname,
|
|
'assettypes': list(kind.assettypes),
|
|
'endpoint': '/api/backups/asset/{assetid}?kind=' + kind.key,
|
|
'render': 'list',
|
|
'map': {
|
|
'title': 'label',
|
|
'meta': [
|
|
{'key': 'collectedat', 'label': 'Captured', 'format': 'date'},
|
|
# Captured is when this config was WRITTEN; verified is
|
|
# when a PC last posted it and the hash still matched.
|
|
# An unchanged config writes no revision, so without
|
|
# this a machine checked daily for a year shows one
|
|
# entry from last spring and reads as abandoned.
|
|
{'key': 'lastseenat', 'label': 'Verified', 'format': 'date'},
|
|
{'key': 'sourcehostname', 'label': 'From'},
|
|
{'key': 'shorthash', 'label': 'Hash', 'mono': True},
|
|
],
|
|
},
|
|
# The generic renderer has no per-item actions, and downloading
|
|
# a .reg needs a per-revision control plus a dialect choice, so
|
|
# the panel links to the plugin's own history view instead.
|
|
'manage': {
|
|
'to': '/backups/asset/{assetid}',
|
|
'label': 'History / download',
|
|
'emptylabel': 'View history',
|
|
},
|
|
'position': 40 + position,
|
|
}
|
|
# OMIT 'empty' entirely unless the kind sets it. The list renderer
|
|
# shows a panel when it has rows OR declares empty text, so a
|
|
# present-but-null 'empty' would still render the panel on every
|
|
# asset of the type. Most machines are not part markers and never
|
|
# will be, so that panel must disappear rather than announce it has
|
|
# nothing.
|
|
if kind.emptytext:
|
|
panel['empty'] = kind.emptytext
|
|
panels.append(panel)
|
|
return panels
|
|
|
|
def get_settings_cards(self) -> List[Dict]:
|
|
"""ADR-010: contribute the Backups settings card.
|
|
|
|
The core settings catalog (settingsNav.js) is a hardcoded list a plugin
|
|
cannot extend, so this plugin shipped five settings and no page for any
|
|
of them - backups_shareroot and the collection interval were reachable
|
|
only through the API or a shell on the server, which is not a
|
|
configurable setting in any useful sense.
|
|
|
|
Filed under Machines because that is where the assets these back up
|
|
live, and it puts the card next to Machine Types rather than in a
|
|
platform group nobody visits.
|
|
"""
|
|
return [
|
|
{
|
|
'group': 'Machines',
|
|
'to': '/settings/backups',
|
|
'icon': 'package',
|
|
'title': 'Backups',
|
|
'description': 'Collection interval, share root for part-marker '
|
|
'snapshots, staleness threshold and revision retention',
|
|
'position': 21,
|
|
},
|
|
]
|
|
|
|
def get_settings_defaults(self) -> List[Dict]:
|
|
"""Schedule + retention, read by the collector and the prune job.
|
|
|
|
The interval lives here rather than in the manifest entry because
|
|
GE-Enforce fires every cycle (several times a day on a shift change).
|
|
The collector reads backups_intervalhours and no-ops when its marker
|
|
file is younger than that, so the cadence is changed centrally in
|
|
ShopDB instead of by editing a script on the share.
|
|
"""
|
|
return [
|
|
{
|
|
'key': 'backups_intervalhours',
|
|
'value': '24',
|
|
'valuetype': 'integer',
|
|
'category': 'backups',
|
|
'description': 'Minimum hours between collection attempts on a '
|
|
'PC. GE-Enforce runs every cycle; the collector '
|
|
'skips until this much time has passed.',
|
|
# MUST be public. The collecting script reads this from
|
|
# /api/settings/public BEFORE it has any credential, and its
|
|
# Get-IntervalHours falls back to 24 on any failure - silently.
|
|
# Left off the allowlist, the setting looked configurable and
|
|
# was not: every PC used 24 no matter what the UI said. A
|
|
# collection cadence is not a secret; the share root beneath it
|
|
# is internal topology and stays private.
|
|
'public': True,
|
|
},
|
|
{
|
|
'key': 'backups_shareroot',
|
|
'value': DEFAULTSHAREROOT,
|
|
'valuetype': 'string',
|
|
'category': 'backups',
|
|
'description': 'UNC root that opaque (non-JSON) backups are '
|
|
'written under by the collecting PC, e.g. '
|
|
'\\\\fileserver\\shopfloor\\backups. Blank until '
|
|
'the site sets it: no default can be right at '
|
|
'more than one site (ADR-015).',
|
|
},
|
|
{
|
|
'key': 'backups_staledays',
|
|
'value': '3',
|
|
'valuetype': 'integer',
|
|
'category': 'backups',
|
|
'description': 'Days without a CONFIRMED check before a backup '
|
|
'is listed as stopped on the dashboard. Measured '
|
|
'from the last check, not the last change - an '
|
|
'unchanged config writes no revision. 0 disables '
|
|
'the card.',
|
|
},
|
|
{
|
|
'key': 'backups_retentioncount',
|
|
'value': '50',
|
|
'valuetype': 'integer',
|
|
'category': 'backups',
|
|
'description': 'Distinct revisions kept per asset per kind. '
|
|
'0 keeps everything. The newest and the oldest '
|
|
'are never pruned.',
|
|
},
|
|
{
|
|
'key': 'backups_retentiondays',
|
|
'value': '0',
|
|
'valuetype': 'integer',
|
|
'category': 'backups',
|
|
'description': 'Also prune revisions older than this many days. '
|
|
'0 disables age-based pruning.',
|
|
},
|
|
]
|
|
|
|
def get_dashboard_widgets(self) -> List[Dict]:
|
|
"""Dashboard card: backups that have stopped running.
|
|
|
|
Keyed on the last CONFIRMED check, never on the last change. Dedup means
|
|
an unchanged config writes no revision, so a card keyed on revision age
|
|
would flag most of a healthy fleet.
|
|
"""
|
|
return [
|
|
{
|
|
'id': 'backups-stale',
|
|
'title': 'Backups stopped',
|
|
'shortname': 'Backups',
|
|
'endpoint': '/api/backups/dashboard/stale',
|
|
'render': 'exceptions',
|
|
'severity': 'warning',
|
|
'permission': 'backups.view',
|
|
'empty': 'hide',
|
|
'position': 30,
|
|
'map': {
|
|
'title': 'assetnumber',
|
|
'detail': 'backupkind',
|
|
'meta': [
|
|
{'key': 'sourcehostname', 'label': 'from'},
|
|
{'key': 'quietdays', 'label': 'last checked',
|
|
'suffix': ' days ago'},
|
|
],
|
|
'link': '/backups/asset/{assetid}',
|
|
},
|
|
},
|
|
]
|
|
|
|
# ---- ADR-006 collector contract -------------------------------------
|
|
|
|
def get_collector_schema(self) -> Optional[dict]:
|
|
return {
|
|
'identityfield': 'machinenumber',
|
|
'fields': {
|
|
'type': 'object',
|
|
'required': ['machinenumber', 'backupkind'],
|
|
'properties': {
|
|
'machinenumber': {
|
|
'type': 'string',
|
|
'description': 'Machine number the config belongs to. '
|
|
'Resolved against assets.assetnumber.',
|
|
},
|
|
'backupkind': {
|
|
'type': 'string',
|
|
'enum': sorted(REGISTRY.keys()),
|
|
},
|
|
'contentbase64': {
|
|
'type': 'string',
|
|
'description': "Raw backup file, base64. Required for "
|
|
"'shopdb' kinds (parsed server-side).",
|
|
},
|
|
'contenthash': {
|
|
'type': 'string',
|
|
'description': "sha256 of the file. Required for 'share' "
|
|
"kinds, whose bytes stay on the SFLD share.",
|
|
},
|
|
'sharepath': {
|
|
'type': 'string',
|
|
'description': "Full UNC path. Required for 'share' kinds.",
|
|
},
|
|
'manifest': {
|
|
'type': 'object',
|
|
'description': "Optional description of a 'share' kind's "
|
|
"contents - what is in the snapshot, not "
|
|
"the bytes themselves. Lets an asset panel "
|
|
"list a snapshot and diff it against the "
|
|
"previous one without the app server "
|
|
"mounting the share.",
|
|
},
|
|
'sourcefilename': {'type': 'string'},
|
|
'sourcehostname': {
|
|
'type': 'string',
|
|
'description': 'PC the backup was read from. For NTLARS '
|
|
'this is the controlling PC, not the machine.',
|
|
},
|
|
'collectedat': {'type': 'string', 'format': 'date-time'},
|
|
'bytesize': {'type': 'integer'},
|
|
},
|
|
},
|
|
}
|
|
|
|
def apply_collector_payload(self, payload: dict) -> dict:
|
|
from shopdb.api import db
|
|
|
|
warnings = []
|
|
kindkey = (payload.get('backupkind') or '').strip().lower()
|
|
kind = getkind(kindkey)
|
|
if kind is None:
|
|
raise ValueError('unknown backupkind: {!r}'.format(kindkey))
|
|
|
|
assetid, err = kind.resolveassetid(payload)
|
|
if assetid is None:
|
|
# Unresolvable is reported, never silently dropped - an unmatched
|
|
# machine number means a real gap between the fleet and ShopDB.
|
|
raise ValueError('could not resolve asset: {}'.format(err))
|
|
|
|
projection = None
|
|
sharepath = None
|
|
sourcefilename = payload.get('sourcefilename')
|
|
bytesize = payload.get('bytesize')
|
|
|
|
if kind.storagebackend == 'shopdb':
|
|
encoded = payload.get('contentbase64')
|
|
if not encoded:
|
|
raise ValueError(
|
|
"contentbase64 is required for kind '{}'".format(kindkey))
|
|
raw = base64.b64decode(encoded)
|
|
bytesize = bytesize or len(raw)
|
|
projection = kind.parse(raw)
|
|
contenthash = canonicalhash(projection)
|
|
|
|
# The fleet's machine number (pc-config.txt) is authoritative for
|
|
# WHICH asset this belongs to, but NTLARS carries its own MachineNo.
|
|
# A disagreement means the PC is running another machine's config -
|
|
# worth surfacing rather than quietly filing under the reported one.
|
|
embedded = getattr(kind, 'embeddedmachineno', None)
|
|
if embedded is not None:
|
|
embeddedno = embedded(projection)
|
|
reported = (payload.get('machinenumber') or '').strip()
|
|
if embeddedno and reported and embeddedno != reported:
|
|
warnings.append(
|
|
'reported machine number {!r} but NTLARS is configured '
|
|
'for {!r}'.format(reported, embeddedno))
|
|
else:
|
|
contenthash = (payload.get('contenthash') or '').strip().lower()
|
|
sharepath = (payload.get('sharepath') or '').strip()
|
|
if not contenthash or not sharepath:
|
|
raise ValueError(
|
|
"contenthash and sharepath are required for kind '{}'".format(
|
|
kindkey))
|
|
|
|
# A share kind may also describe itself. The bytes stay on the
|
|
# share - that part is unchanged - but a manifest of what the
|
|
# snapshot CONTAINS is small, structured, and the only way an asset
|
|
# panel can list a part marker's patterns or say which one
|
|
# disappeared last month. The alternative is the app server
|
|
# mounting SFLD, which the read routes deliberately refuse to do.
|
|
#
|
|
# It is advisory: dedup still keys on contenthash, so a client that
|
|
# sends no manifest behaves exactly as before.
|
|
manifest = payload.get('manifest')
|
|
if manifest is not None:
|
|
# Validated at the door, not at render time. A shallow check
|
|
# here would let a malformed manifest become the LATEST
|
|
# revision, and every view of the asset panel would then fail
|
|
# until the content next changed - a client bug turning into a
|
|
# broken page that outlives it.
|
|
if not isinstance(manifest, dict):
|
|
raise ValueError('manifest must be an object')
|
|
files = manifest.get('files')
|
|
if files is not None:
|
|
if not isinstance(files, list):
|
|
raise ValueError('manifest files must be a list')
|
|
for entry in files:
|
|
if not isinstance(entry, dict):
|
|
raise ValueError(
|
|
'each manifest file must be an object')
|
|
if not isinstance(entry.get('name'), str) or not entry['name']:
|
|
raise ValueError(
|
|
'each manifest file needs a name')
|
|
size = entry.get('bytes')
|
|
if size is not None and not isinstance(size, int):
|
|
raise ValueError(
|
|
'manifest file {!r} has a non-integer size'.format(
|
|
entry['name']))
|
|
projection = manifest
|
|
|
|
# Dedup is against the LATEST row of THIS SOURCE's chain, not the
|
|
# asset's. Several PCs legitimately share one machine number here - the
|
|
# part markers on 0613, 0615 and WJPRT do, and their configs genuinely
|
|
# differ, typically by COM port. Keyed on
|
|
# asset alone, each marker's post differed from whichever marker posted
|
|
# last, so nothing ever deduped and the chain grew by one revision per
|
|
# PC per collection cycle. Keyed on the source, an unchanged config is a
|
|
# no-op again and each marker keeps its own history against the machine.
|
|
#
|
|
# Dedup is against the latest row only, not any row, because a config
|
|
# reverting to an earlier state is a legitimate new revision.
|
|
#
|
|
# with_for_update serializes concurrent posts for the same chain: the
|
|
# fleet collects on one GE-Enforce cycle, so a retry could otherwise
|
|
# pass the hash check twice and insert the same revision.
|
|
# Normalize ONCE, and use the same value to query and to store.
|
|
# samesource() maps '' to IS NULL, but the insert below used to store
|
|
# the raw '' - so a client posting an empty hostname wrote rows that its
|
|
# own next post could never match. Every post then deduped against
|
|
# nothing, inserting forever: unbounded growth, and a history no
|
|
# technician can read. Whitespace does the same thing more quietly.
|
|
sourcehostname = (payload.get('sourcehostname') or '').strip() or None
|
|
latest = (db.session.query(BackupRevision)
|
|
.filter(BackupRevision.assetid == assetid,
|
|
BackupRevision.backupkind == kindkey,
|
|
_samesource(sourcehostname))
|
|
.order_by(BackupRevision.backuprevisionid.desc())
|
|
.with_for_update()
|
|
.first())
|
|
|
|
if latest is not None and latest.contenthash == contenthash:
|
|
# Unchanged, so no revision - but RECORD THE CHECK. Without this the
|
|
# only evidence a backup still runs is a line in a log on the PC,
|
|
# and a config stable for six months is indistinguishable from a
|
|
# backup that died six months ago. Cheap: one column on a row that
|
|
# already exists, no new history.
|
|
latest.lastseenat = datetime.utcnow()
|
|
|
|
# BACKFILL, on a no-op. The contenthash attests the BYTES, not the
|
|
# manifest, so a client that starts describing itself posts the same
|
|
# hash as before and would otherwise no-op forever - leaving the
|
|
# index unstored on exactly the markers that are healthy and
|
|
# unchanging, which is most of them. A revision that already has a
|
|
# manifest is left alone; this only fills a hole.
|
|
#
|
|
# Honest by construction: the hash proves the content is identical,
|
|
# so indexing it now describes the same snapshot. lastseenat
|
|
# semantics are untouched.
|
|
if projection is not None and latest.payload is None:
|
|
latest.payload = projection
|
|
# Same reasoning for the pointer: a snapshot that moved on the share
|
|
# (a changed layout, a renamed asset directory) reports a new
|
|
# sharepath with unchanged bytes, and the stored path would
|
|
# otherwise go on pointing at a directory nobody writes to.
|
|
if sharepath and latest.sharepath != sharepath:
|
|
latest.sharepath = sharepath
|
|
|
|
db.session.flush()
|
|
return {
|
|
'action': 'noop',
|
|
'assetid': assetid,
|
|
'backuprevisionid': latest.backuprevisionid,
|
|
'warnings': warnings,
|
|
}
|
|
|
|
collectedat = None
|
|
rawcollected = payload.get('collectedat')
|
|
if rawcollected:
|
|
try:
|
|
collectedat = datetime.fromisoformat(rawcollected.replace('Z', '+00:00'))
|
|
if collectedat.tzinfo is not None:
|
|
# Convert, don't just drop the offset: createdat is utcnow(),
|
|
# so keeping local wall time would skew a -04:00 bay by 4h
|
|
# against its own history.
|
|
collectedat = collectedat.astimezone(timezone.utc).replace(
|
|
tzinfo=None)
|
|
except ValueError:
|
|
warnings.append('unparseable collectedat: {!r}'.format(rawcollected))
|
|
|
|
revision = BackupRevision(
|
|
assetid=assetid,
|
|
backupkind=kindkey,
|
|
storagebackend=kind.storagebackend,
|
|
contenthash=contenthash,
|
|
sharepath=sharepath,
|
|
sourcefilename=sourcefilename,
|
|
bytesize=bytesize,
|
|
sourcehostname=sourcehostname,
|
|
collectedat=collectedat or datetime.utcnow(),
|
|
lastseenat=datetime.utcnow(),
|
|
)
|
|
revision.payload = projection
|
|
db.session.add(revision)
|
|
# flush, not commit: the collector dispatcher owns the transaction and
|
|
# commits after writing its AuditLog row. Committing here would leave an
|
|
# unaudited revision behind if that write then failed.
|
|
db.session.flush()
|
|
|
|
pruned = prune(
|
|
assetid, kindkey,
|
|
# Setting.get, NOT self.get_setting. The BasePlugin helper
|
|
# namespaces what it reads to `plugin.backups.<key>`, while
|
|
# get_settings_defaults declares - and the settings page writes -
|
|
# the bare key. So this read never found the value an operator had
|
|
# set, always fell back to 0, and 0 means "keep everything": the
|
|
# retention policy was configurable in the UI and did nothing.
|
|
retentioncount=Setting.get('backups_retentioncount', 0),
|
|
retentiondays=Setting.get('backups_retentiondays', 0),
|
|
)
|
|
if pruned:
|
|
warnings.append('pruned {} old revision(s) per retention policy'
|
|
.format(pruned))
|
|
|
|
return {
|
|
'action': 'created',
|
|
'assetid': assetid,
|
|
'backuprevisionid': revision.backuprevisionid,
|
|
'warnings': warnings,
|
|
}
|
|
|
|
def init_app(self, app: Flask, db_instance) -> None:
|
|
logger.info('Backups plugin initialized (v%s, kinds: %s)',
|
|
self.meta.version, ', '.join(sorted(REGISTRY)))
|