fix(fleet-sync): make control-identity-mismatch sticky and surface in UI (#1117)

Treat 409 CONTROL_IDENTITY_MISMATCH from a replica as a non-retriable
failure instead of looping the same 409 through the 5-minute retry
service forever and silently writing identical failure rows.

Backend
- DatabaseService: add `sticky_error_code`, `sticky_error_expected`,
  `sticky_error_got` columns to `fleet_sync_status` via an idempotent
  migration. New methods setFleetSyncSticky, getFleetSyncStickyCode,
  clearFleetSyncStickyForNode. recordFleetSyncSuccess clears the sticky
  flag on a clean push. getFailedSyncTargets SQL adds
  `AND sticky_error_code IS NULL` so the retry loop skips sticky rows.
- FleetSyncService.executePushToNode: short-circuits at the top when
  sticky is set (covers event-driven pushResourceAsync calls). On a 409
  with code CONTROL_IDENTITY_MISMATCH, records the failure once and
  pins sticky with the expected/got fingerprints carried in the 409 body.
- routes/nodes.ts: new POST /api/nodes/:id/fleet-sync/reset-anchor.
  Admin + paid + node:manage. Proxies POST /api/fleet/role/reanchor to
  the peer with `{override:true}` using the stored Bearer node_proxy
  token. On peer 200, clears every sticky row for the node so the next
  push re-anchors and resumes replication. Distinct 502 / 504 responses
  for peer-rejected / peer-unreachable so the UI can show a useful toast.

Frontend
- New lib/fleetSyncApi.ts + hooks/useFleetSyncStatus.ts. Polling hook
  (30s visibilityInterval) skips fetch when !isPaid.
- NodeManager.tsx: destructive banner per affected node listing both
  fingerprints, with `Reset anchor on peer` and `Remove node` buttons.
  Hidden for community-tier users via empty hook data.
- FleetConfiguration.tsx (Fleet -> Status): read-only `Policy sync`
  SummaryRow per remote node card. In sync / degraded / paused with
  a tooltip; no action buttons (the action lives in NodeManager).

Tests
- fleet-sync-service.test.ts: 4 new cases for sticky-set on first
  mismatch, short-circuit on subsequent pushes, null fingerprints,
  and non-mismatch failures not setting sticky.
- database-fleet-sync-sticky.test.ts (new): 6 cases pinning the DB
  contract incl. retry-loop SQL filter and migration idempotency.
- nodes-fleet-sync-reset-anchor.test.ts (new): 6 cases covering
  happy path, peer 401 -> 502, peer unreachable -> 504, local-node
  rejection, unknown node id, and community-tier 403.

Gate parity (Directive 30): the new POST .../reset-anchor enforces
requireAdmin + requirePaid + node:manage (matches the existing read at
GET /api/fleet/sync-status). UI banner + SummaryRow only render when the
hook returns data, which it only does for paid-tier authed users. No
existing tier-gate file moved; this is greenfield parity.

Auth audit: the peer's POST /api/fleet/role/reanchor route already uses
requireAdmin, which accepts the central's stored node_proxy Bearer
token because authMiddleware maps `scope === 'node_proxy'` to
`req.user = { username: 'node-proxy', role: 'admin', userId: 0 }`.
No widening required.

Backend tsc clean. Frontend tsc -b clean. 59 fleet-sync tests pass; full
backend suite green minus the pre-existing Windows-only file-lock flake
on filesystem-backup.test.ts that reproduces unchanged on main.
This commit is contained in:
Anso
2026-05-19 19:49:16 -04:00
committed by GitHub
parent 69bc955c3b
commit e05099f2a1
10 changed files with 876 additions and 11 deletions
@@ -1,12 +1,15 @@
import { useState, useEffect, useCallback } from 'react';
import { useState, useEffect, useCallback, useMemo } from 'react';
import { apiFetch } from '@/lib/api';
import { formatCount } from '@/lib/utils';
import { Card, CardContent } from '@/components/ui/card';
import { Badge } from '@/components/ui/badge';
import { Tooltip, TooltipContent, TooltipProvider, TooltipTrigger } from '@/components/ui/tooltip';
import {
Bell, Zap, Shield, HardDrive, WifiOff, CheckCircle2,
Bell, Zap, Shield, HardDrive, WifiOff, CheckCircle2, RefreshCw,
} from 'lucide-react';
import { useLicense } from '@/context/LicenseContext';
import { useFleetSyncStatus } from '@/hooks/useFleetSyncStatus';
import { STICKY_CONTROL_IDENTITY_MISMATCH, type FleetSyncStatus } from '@/lib/fleetSyncApi';
import type { ConfigurationStatusPayload } from '@/components/dashboard';
interface FleetNodeConfiguration {
@@ -31,7 +34,69 @@ function SummaryRow({ icon: Icon, label, value }: {
);
}
function NodeCard({ node, isPaid }: { node: FleetNodeConfiguration; isPaid: boolean }) {
type PolicySyncState =
| { kind: 'in_sync' }
| { kind: 'degraded'; lastError: string | null }
| { kind: 'paused' };
function derivePolicySyncState(rows: FleetSyncStatus[]): PolicySyncState | null {
if (rows.length === 0) return null;
for (const row of rows) {
if (row.sticky_error_code === STICKY_CONTROL_IDENTITY_MISMATCH) {
return { kind: 'paused' };
}
}
let degradedError: string | null = null;
let hasSuccess = false;
for (const row of rows) {
if (row.last_success_at !== null) hasSuccess = true;
if (
row.last_failure_at !== null
&& (row.last_success_at === null || row.last_failure_at > row.last_success_at)
) {
degradedError = row.last_error;
}
}
if (degradedError !== null) return { kind: 'degraded', lastError: degradedError };
if (hasSuccess) return { kind: 'in_sync' };
return null;
}
function PolicySyncRow({ state }: { state: PolicySyncState }) {
if (state.kind === 'in_sync') {
return <SummaryRow icon={RefreshCw} label="Policy sync" value="In sync" />;
}
const tooltip = state.kind === 'paused'
? 'Anchored to another central. Open Settings → Nodes to reset the anchor or remove the node.'
: (state.lastError ?? 'Last push to this node failed.');
return (
<div className="flex items-center gap-2 py-0.5">
<RefreshCw className="h-3 w-3 shrink-0 text-stat-icon" strokeWidth={1.5} />
<span className="text-xs flex-1 text-stat-subtitle">Policy sync</span>
<TooltipProvider>
<Tooltip>
<TooltipTrigger>
<span
className={
'inline-flex items-center rounded border px-1.5 py-0 text-[10px] leading-3 font-mono uppercase '
+ 'border-amber-500/40 bg-amber-500/10 text-amber-600 dark:text-amber-400'
}
>
{state.kind === 'paused' ? 'paused' : 'degraded'}
</span>
</TooltipTrigger>
<TooltipContent>{tooltip}</TooltipContent>
</Tooltip>
</TooltipProvider>
</div>
);
}
function NodeCard({ node, isPaid, policySyncState }: {
node: FleetNodeConfiguration;
isPaid: boolean;
policySyncState: PolicySyncState | null;
}) {
const isRemote = node.type === 'remote';
if (!node.configuration) {
return (
@@ -102,6 +167,7 @@ function NodeCard({ node, isPaid }: { node: FleetNodeConfiguration; isPaid: bool
)}
<SummaryRow icon={HardDrive} label="Crash detect"
value={thresholds.globalCrash ? 'On' : 'Off'} />
{policySyncState && <PolicySyncRow state={policySyncState} />}
</div>
</CardContent>
</Card>
@@ -110,10 +176,26 @@ function NodeCard({ node, isPaid }: { node: FleetNodeConfiguration; isPaid: bool
export function FleetConfiguration() {
const { isPaid } = useLicense();
const { statuses: syncStatuses } = useFleetSyncStatus();
const [nodes, setNodes] = useState<FleetNodeConfiguration[]>([]);
const [loading, setLoading] = useState(true);
const [error, setError] = useState<string | null>(null);
const syncStateByNode = useMemo(() => {
const byNode = new Map<number, FleetSyncStatus[]>();
for (const row of syncStatuses) {
const list = byNode.get(row.node_id);
if (list) list.push(row);
else byNode.set(row.node_id, [row]);
}
const out = new Map<number, PolicySyncState>();
for (const [nodeId, rows] of byNode) {
const state = derivePolicySyncState(rows);
if (state) out.set(nodeId, state);
}
return out;
}, [syncStatuses]);
const fetchData = useCallback(async () => {
try {
const res = await apiFetch('/fleet/configuration', { localOnly: true });
@@ -170,7 +252,14 @@ export function FleetConfiguration() {
return (
<div className="grid grid-cols-1 md:grid-cols-2 xl:grid-cols-3 gap-4 p-1">
{nodes.map(node => <NodeCard key={node.id} node={node} isPaid={isPaid} />)}
{nodes.map(node => (
<NodeCard
key={node.id}
node={node}
isPaid={isPaid}
policySyncState={syncStateByNode.get(node.id) ?? null}
/>
))}
</div>
);
}