fix(federation): use node http healthcheck and any-restart for gateway test stack

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

apps/gateway/src/federation/scope-schema.spec.ts

@@ -0,0 +1,187 @@
+/**
+ * Unit tests for FederationScopeSchema and parseFederationScope.
+ *
+ * Coverage:
+ *  - Valid: minimal scope
+ *  - Valid: full PRD §8.1 example
+ *  - Valid: resources + excluded_resources (no overlap)
+ *  - Invalid: empty resources
+ *  - Invalid: unknown resource value
+ *  - Invalid: resources / excluded_resources intersection
+ *  - Invalid: filter key not in resources
+ *  - Invalid: max_rows_per_query = 0
+ *  - Invalid: max_rows_per_query = 10001
+ *  - Invalid: not an object / null
+ *  - Defaults: include_personal defaults to true; excluded_resources defaults to []
+ *  - Sentinel: console.warn fires for sensitive resources
+ */
+
+import { describe, it, expect, vi, afterEach } from 'vitest';
+import {
+  parseFederationScope,
+  FederationScopeError,
+  FederationScopeSchema,
+} from './scope-schema.js';
+
+afterEach(() => {
+  vi.restoreAllMocks();
+});
+
+describe('parseFederationScope — valid inputs', () => {
+  it('accepts a minimal scope (resources + max_rows_per_query only)', () => {
+    const scope = parseFederationScope({
+      resources: ['tasks'],
+      max_rows_per_query: 100,
+    });
+    expect(scope.resources).toEqual(['tasks']);
+    expect(scope.max_rows_per_query).toBe(100);
+    expect(scope.excluded_resources).toEqual([]);
+    expect(scope.filters).toBeUndefined();
+  });
+
+  it('accepts the full PRD §8.1 example', () => {
+    const scope = parseFederationScope({
+      resources: ['tasks', 'notes', 'memory'],
+      filters: {
+        tasks: { include_teams: ['team_uuid_1', 'team_uuid_2'], include_personal: true },
+        notes: { include_personal: true, include_teams: [] },
+        memory: { include_personal: true },
+      },
+      excluded_resources: ['credentials', 'api_keys'],
+      max_rows_per_query: 500,
+    });
+    expect(scope.resources).toEqual(['tasks', 'notes', 'memory']);
+    expect(scope.excluded_resources).toEqual(['credentials', 'api_keys']);
+    expect(scope.filters?.tasks?.include_teams).toEqual(['team_uuid_1', 'team_uuid_2']);
+    expect(scope.max_rows_per_query).toBe(500);
+  });
+
+  it('accepts a scope with excluded_resources and no filter overlap', () => {
+    const scope = parseFederationScope({
+      resources: ['tasks', 'notes'],
+      excluded_resources: ['memory'],
+      max_rows_per_query: 250,
+    });
+    expect(scope.resources).toEqual(['tasks', 'notes']);
+    expect(scope.excluded_resources).toEqual(['memory']);
+  });
+});
+
+describe('parseFederationScope — defaults', () => {
+  it('defaults excluded_resources to []', () => {
+    const scope = parseFederationScope({ resources: ['tasks'], max_rows_per_query: 1 });
+    expect(scope.excluded_resources).toEqual([]);
+  });
+
+  it('defaults include_personal to true when filter is provided without it', () => {
+    const scope = parseFederationScope({
+      resources: ['tasks'],
+      filters: { tasks: { include_teams: ['t1'] } },
+      max_rows_per_query: 10,
+    });
+    expect(scope.filters?.tasks?.include_personal).toBe(true);
+  });
+});
+
+describe('parseFederationScope — invalid inputs', () => {
+  it('throws FederationScopeError for empty resources array', () => {
+    expect(() => parseFederationScope({ resources: [], max_rows_per_query: 100 })).toThrow(
+      FederationScopeError,
+    );
+  });
+
+  it('throws for unknown resource value in resources', () => {
+    expect(() =>
+      parseFederationScope({ resources: ['unknown_resource'], max_rows_per_query: 100 }),
+    ).toThrow(FederationScopeError);
+  });
+
+  it('throws when resources and excluded_resources intersect', () => {
+    expect(() =>
+      parseFederationScope({
+        resources: ['tasks', 'memory'],
+        excluded_resources: ['memory'],
+        max_rows_per_query: 100,
+      }),
+    ).toThrow(FederationScopeError);
+  });
+
+  it('throws when filters references a resource not in resources', () => {
+    expect(() =>
+      parseFederationScope({
+        resources: ['tasks'],
+        filters: { notes: { include_personal: true } },
+        max_rows_per_query: 100,
+      }),
+    ).toThrow(FederationScopeError);
+  });
+
+  it('throws for max_rows_per_query = 0', () => {
+    expect(() => parseFederationScope({ resources: ['tasks'], max_rows_per_query: 0 })).toThrow(
+      FederationScopeError,
+    );
+  });
+
+  it('throws for max_rows_per_query = 10001', () => {
+    expect(() => parseFederationScope({ resources: ['tasks'], max_rows_per_query: 10001 })).toThrow(
+      FederationScopeError,
+    );
+  });
+
+  it('throws for null input', () => {
+    expect(() => parseFederationScope(null)).toThrow(FederationScopeError);
+  });
+
+  it('throws for non-object input (string)', () => {
+    expect(() => parseFederationScope('not-an-object')).toThrow(FederationScopeError);
+  });
+});
+
+describe('parseFederationScope — sentinel warning', () => {
+  it('emits console.warn when resources includes "credentials"', () => {
+    const warnSpy = vi.spyOn(console, 'warn').mockImplementation(() => {});
+    parseFederationScope({
+      resources: ['tasks', 'credentials'],
+      max_rows_per_query: 100,
+    });
+    expect(warnSpy).toHaveBeenCalledWith(
+      expect.stringContaining(
+        '[FederationScope] WARNING: scope grants sensitive resource "credentials"',
+      ),
+    );
+  });
+
+  it('emits console.warn when resources includes "api_keys"', () => {
+    const warnSpy = vi.spyOn(console, 'warn').mockImplementation(() => {});
+    parseFederationScope({
+      resources: ['tasks', 'api_keys'],
+      max_rows_per_query: 100,
+    });
+    expect(warnSpy).toHaveBeenCalledWith(
+      expect.stringContaining(
+        '[FederationScope] WARNING: scope grants sensitive resource "api_keys"',
+      ),
+    );
+  });
+
+  it('does NOT emit console.warn for non-sensitive resources', () => {
+    const warnSpy = vi.spyOn(console, 'warn').mockImplementation(() => {});
+    parseFederationScope({ resources: ['tasks', 'notes', 'memory'], max_rows_per_query: 100 });
+    expect(warnSpy).not.toHaveBeenCalled();
+  });
+});
+
+describe('FederationScopeSchema — boundary values', () => {
+  it('accepts max_rows_per_query = 1 (lower bound)', () => {
+    const result = FederationScopeSchema.safeParse({ resources: ['tasks'], max_rows_per_query: 1 });
+    expect(result.success).toBe(true);
+  });
+
+  it('accepts max_rows_per_query = 10000 (upper bound)', () => {
+    const result = FederationScopeSchema.safeParse({
+      resources: ['tasks'],
+      max_rows_per_query: 10000,
+    });
+    expect(result.success).toBe(true);
+  });
+});

apps/gateway/src/federation/scope-schema.ts

@@ -0,0 +1,147 @@
+/**
+ * Federation grant scope schema and validator.
+ *
+ * Source of truth: docs/federation/PRD.md §8.1
+ *
+ * This module is intentionally pure — no DB, no NestJS, no CA wiring.
+ * It is reusable from grant CRUD (M2-06) and scope enforcement (M3+).
+ */
+
+import { z } from 'zod';
+
+// ---------------------------------------------------------------------------
+// Allowlist of federation resources (canonical — M3+ will extend this list)
+// ---------------------------------------------------------------------------
+export const FEDERATION_RESOURCE_VALUES = [
+  'tasks',
+  'notes',
+  'memory',
+  'credentials',
+  'api_keys',
+] as const;
+
+export type FederationResource = (typeof FEDERATION_RESOURCE_VALUES)[number];
+
+/**
+ * Sensitive resources require explicit admin approval (PRD §8.4).
+ * The parser warns when these appear in `resources`; M2-06 grant CRUD
+ * will add a hard gate on top of this warning.
+ */
+const SENSITIVE_RESOURCES: ReadonlySet<FederationResource> = new Set(['credentials', 'api_keys']);
+
+// ---------------------------------------------------------------------------
+// Sub-schemas
+// ---------------------------------------------------------------------------
+
+const ResourceArraySchema = z
+  .array(z.enum(FEDERATION_RESOURCE_VALUES))
+  .nonempty({ message: 'resources must contain at least one value' })
+  .refine((arr) => new Set(arr).size === arr.length, {
+    message: 'resources must not contain duplicate values',
+  });
+
+const ResourceFilterSchema = z.object({
+  include_teams: z.array(z.string()).optional(),
+  include_personal: z.boolean().default(true),
+});
+
+// ---------------------------------------------------------------------------
+// Top-level schema
+// ---------------------------------------------------------------------------
+
+export const FederationScopeSchema = z
+  .object({
+    resources: ResourceArraySchema,
+
+    excluded_resources: z
+      .array(z.enum(FEDERATION_RESOURCE_VALUES))
+      .default([])
+      .refine((arr) => new Set(arr).size === arr.length, {
+        message: 'excluded_resources must not contain duplicate values',
+      }),
+
+    filters: z.record(z.string(), ResourceFilterSchema).optional(),
+
+    max_rows_per_query: z
+      .number()
+      .int({ message: 'max_rows_per_query must be an integer' })
+      .min(1, { message: 'max_rows_per_query must be at least 1' })
+      .max(10000, { message: 'max_rows_per_query must be at most 10000' }),
+  })
+  .superRefine((data, ctx) => {
+    const resourceSet = new Set(data.resources);
+
+    // Intersection guard: a resource cannot be both granted and excluded
+    for (const r of data.excluded_resources) {
+      if (resourceSet.has(r)) {
+        ctx.addIssue({
+          code: z.ZodIssueCode.custom,
+          message: `Resource "${r}" appears in both resources and excluded_resources`,
+          path: ['excluded_resources'],
+        });
+      }
+    }
+
+    // Filter keys must be a subset of resources
+    if (data.filters) {
+      for (const key of Object.keys(data.filters)) {
+        if (!resourceSet.has(key as FederationResource)) {
+          ctx.addIssue({
+            code: z.ZodIssueCode.custom,
+            message: `filters key "${key}" references a resource not present in resources`,
+            path: ['filters', key],
+          });
+        }
+      }
+    }
+  });
+
+export type FederationScope = z.infer<typeof FederationScopeSchema>;
+
+// ---------------------------------------------------------------------------
+// Error class
+// ---------------------------------------------------------------------------
+
+export class FederationScopeError extends Error {
+  constructor(message: string) {
+    super(message);
+    this.name = 'FederationScopeError';
+  }
+}
+
+// ---------------------------------------------------------------------------
+// Typed parser
+// ---------------------------------------------------------------------------
+
+/**
+ * Parse and validate an unknown value as a FederationScope.
+ *
+ * Throws `FederationScopeError` with aggregated Zod issues on failure.
+ *
+ * Emits `console.warn` when sensitive resources (`credentials`, `api_keys`)
+ * are present in `resources` — per PRD §8.4, these require explicit admin
+ * approval. M2-06 grant CRUD will add a hard gate on top of this warning.
+ */
+export function parseFederationScope(input: unknown): FederationScope {
+  const result = FederationScopeSchema.safeParse(input);
+
+  if (!result.success) {
+    const issues = result.error.issues
+      .map((e) => `  - [${e.path.join('.') || 'root'}] ${e.message}`)
+      .join('\n');
+    throw new FederationScopeError(`Invalid federation scope:\n${issues}`);
+  }
+
+  const scope = result.data;
+
+  // Sentinel warning for sensitive resources (PRD §8.4)
+  for (const resource of scope.resources) {
+    if (SENSITIVE_RESOURCES.has(resource)) {
+      console.warn(
+        `[FederationScope] WARNING: scope grants sensitive resource "${resource}". Per PRD §8.4 this requires explicit admin approval and is logged.`,
+      );
+    }
+  }
+
+  return scope;
+}

deploy/portainer/federated-test.stack.yml

@@ -30,9 +30,18 @@
 #   DNS A record        ${HOST_FQDN} → Swarm ingress IP (or Cloudflare proxy).
 #
 # IMAGE
-#   Pinned to digest fed-v0.1.0-m1 (DEPLOY-01 verified).
+#   Pinned to sha-9f1a081 (main HEAD post-#488 Dockerfile fix). The previous
+#   pin (fed-v0.1.0-m1, sha256:9b72e2...) had a broken pnpm copy and could
+#   not resolve @mosaicstack/storage at runtime. The new digest was smoke-
+#   tested locally — gateway boots, imports resolve, tier-detector runs.
 #   Update digest here when promoting a new build.
 #
+# HEALTHCHECK NOTE (2026-04-21)
+#   Switched from busybox wget to node http.get on 127.0.0.1 (not localhost) to
+#   avoid IPv6 resolution issues on Alpine. Retries increased to 5 and
+#   start_period to 60s to cover the NestJS/GC cold-start window (~40-50s).
+#   restart_policy set to `any` so SIGTERM/clean-exit also triggers restart.
+#
 # NOTE: This is a TEST template — production deployments use a separate
 #       parameterised template with stricter resource limits and secrets.
 
@@ -40,8 +49,8 @@ version: '3.9'
 
 services:
   gateway:
-    image: git.mosaicstack.dev/mosaicstack/stack/gateway@sha256:9b72e202a9eecc27d31920b87b475b9e96e483c0323acc57856be4b1355db1ec
-    # Tag for human reference: fed-v0.1.0-m1
+    image: git.mosaicstack.dev/mosaicstack/stack/gateway@sha256:1069117740e00ccfeba357cae38c43f3729fe5ae702740ce474f6512414d7c02
+    # Tag for human reference: sha-9f1a081 (post-#488 Dockerfile fix; smoke-tested locally)
     environment:
       # ── Tier ───────────────────────────────────────────────────────────────
       MOSAIC_TIER: federated
@@ -73,7 +82,7 @@ services:
     deploy:
       replicas: 1
       restart_policy:
-        condition: on-failure
+        condition: any
         delay: 5s
         max_attempts: 3
       labels:
@@ -85,11 +94,15 @@ services:
         - 'traefik.http.routers.${STACK_NAME}.tls.certresolver=letsencrypt'
         - 'traefik.http.services.${STACK_NAME}.loadbalancer.server.port=3000'
     healthcheck:
-      test: ['CMD', 'wget', '-qO-', 'http://localhost:3000/health']
+      test:
+        - 'CMD'
+        - 'node'
+        - '-e'
+        - "require('http').get('http://127.0.0.1:3000/health',r=>process.exit(r.statusCode===200?0:1)).on('error',()=>process.exit(1))"
       interval: 30s
       timeout: 5s
-      retries: 3
-      start_period: 20s
+      retries: 5
+      start_period: 60s
     depends_on:
       - postgres
       - valkey