[management] resolve private services on custom domains in synthesized DNS zones (#6348)

private services on a custom domain didn't resolve on clients — the synthesized DNS zone was anchored to the cluster, and the account's custom domains weren't even
  loaded.

- account.go — SynthesizePrivateServiceZones now keys zones by a resolved apex (privateServiceDomainZone): cluster suffix → registered account.Domains (filtered by matching
  TargetCluster, longest wins) → skip if none. One zone per apex; custom-domain services group under their registered domain.
- sql_store.go — GetAccount now loads account.Domains on both loaders (gorm Preload("Domains") + pgx goroutine via ListCustomDomains; errChan buffer bumped 12→16). This was
  the reason the deploy didn't work — the relation was empty in prod.
- Tests — custom-domain zone synthesis cases (apex resolution, free+custom separation, sibling collapse, cluster mismatch, mixed cluster/custom/public) + GetAccount
  domain-preload tests on sqlite and Postgres.
This commit is contained in:
Maycon Santos
2026-06-06 12:56:01 +02:00
committed by GitHub
parent b377d99933
commit 1e7b16db0a
5 changed files with 290 additions and 10 deletions
+13 -1
View File
@@ -1216,6 +1216,7 @@ func (s *SqlStore) getAccountGorm(ctx context.Context, accountID string) (*types
Preload("NetworkResources").
Preload("Onboarding").
Preload("Services.Targets").
Preload("Domains").
Take(&account, idQueryCondition, accountID)
if result.Error != nil {
log.WithContext(ctx).Errorf("error when getting account %s from the store: %s", accountID, result.Error)
@@ -1302,7 +1303,7 @@ func (s *SqlStore) getAccountPgx(ctx context.Context, accountID string) (*types.
}
var wg sync.WaitGroup
errChan := make(chan error, 12)
errChan := make(chan error, 16)
wg.Add(1)
go func() {
@@ -1403,6 +1404,17 @@ func (s *SqlStore) getAccountPgx(ctx context.Context, accountID string) (*types.
account.Services = services
}()
wg.Add(1)
go func() {
defer wg.Done()
domains, err := s.ListCustomDomains(ctx, accountID)
if err != nil {
errChan <- err
return
}
account.Domains = domains
}()
wg.Add(1)
go func() {
defer wg.Done()
@@ -4,6 +4,8 @@ import (
"context"
"net"
"net/netip"
"os"
"runtime"
"testing"
"time"
@@ -21,6 +23,63 @@ import (
"github.com/netbirdio/netbird/route"
)
// TestGetAccount_LoadsCustomDomains verifies GetAccount populates account.Domains.
// SynthesizePrivateServiceZones depends on this relation to anchor a custom-domain
// private service's DNS zone; without the preload the relation is empty and the
// service is silently skipped, so a custom domain never resolves on clients.
func TestGetAccount_LoadsCustomDomains(t *testing.T) {
if runtime.GOOS == "windows" {
t.Skip("The SQLite store is not properly supported by Windows yet")
}
store, cleanup, err := NewTestStoreFromSQL(context.Background(), "", t.TempDir())
require.NoError(t, err)
defer cleanup()
assertGetAccountLoadsCustomDomains(t, store)
}
func TestPostgresql_GetAccount_LoadsCustomDomains(t *testing.T) {
if (os.Getenv("CI") == "true" && runtime.GOOS == "darwin") || runtime.GOOS == "windows" {
t.Skip("skip CI tests on darwin and windows")
}
t.Setenv("NETBIRD_STORE_ENGINE", string(types.PostgresStoreEngine))
store, cleanup, err := NewTestStoreFromSQL(context.Background(), "", t.TempDir())
require.NoError(t, err)
t.Cleanup(cleanup)
assertGetAccountLoadsCustomDomains(t, store)
}
// assertGetAccountLoadsCustomDomains exercises both the gorm and pgx GetAccount
// paths: it persists two custom domains and asserts the relation comes back
// populated, which SynthesizePrivateServiceZones relies on.
func assertGetAccountLoadsCustomDomains(t *testing.T, store Store) {
t.Helper()
ctx := context.Background()
accountID := "acct-custom-domains"
require.NoError(t, store.SaveAccount(ctx, newAccountWithId(ctx, accountID, "user-1", "")))
_, err := store.CreateCustomDomain(ctx, accountID, "example.com", "eu.proxy.netbird.io", true)
require.NoError(t, err, "creating the first custom domain must succeed")
_, err = store.CreateCustomDomain(ctx, accountID, "apps.acme.io", "us.proxy.netbird.io", false)
require.NoError(t, err, "creating the second custom domain must succeed")
account, err := store.GetAccount(ctx, accountID)
require.NoError(t, err)
require.Len(t, account.Domains, 2, "GetAccount must preload the account's custom domains")
byDomain := map[string]string{}
for _, d := range account.Domains {
require.NotNil(t, d)
byDomain[d.Domain] = d.TargetCluster
}
assert.Equal(t, "eu.proxy.netbird.io", byDomain["example.com"], "custom domain must carry its target cluster")
assert.Equal(t, "us.proxy.netbird.io", byDomain["apps.acme.io"], "custom domain must carry its target cluster")
}
// TestGetAccount_ComprehensiveFieldValidation validates that GetAccount properly loads
// all fields and nested objects from the database, including deeply nested structures.
func TestGetAccount_ComprehensiveFieldValidation(t *testing.T) {