fix(cache): prevent org cache overwrite by other instances (#10012)

# Which Problems Are Solved

A customer reported that randomly certain login flows, such as automatic
redirect to the only configured IdP would not work. During the
investigation it was discovered that they used that same primary domain
on two different instances. As they used the domain for preselecting the
organization, one would always overwrite the other in the cache. Since
The organization and especially it's policies could not be retrieved on
the other instance, it would fallback to the default organization
settings, where the external login and the corresponding IdP were not
configured.

# How the Problems Are Solved

Include the instance id in the cache key for organizations to prevent
overwrites.

# Additional Changes

None

# Additional Context

- found because of a support request
- requires backport to 2.70.x, 2.71.x and 3.x

(cherry picked from commit 15902f5bc7)

internal/query/org.go

@@ -12,6 +12,7 @@ import (
 
 	"github.com/zitadel/zitadel/internal/api/authz"
 	domain_pkg "github.com/zitadel/zitadel/internal/domain"
+	es "github.com/zitadel/zitadel/internal/eventstore"
 	"github.com/zitadel/zitadel/internal/eventstore/handler/v2"
 	"github.com/zitadel/zitadel/internal/feature"
 	"github.com/zitadel/zitadel/internal/query/projection"
@@ -77,6 +78,8 @@ type Org struct {
 	ResourceOwner string
 	State         domain_pkg.OrgState
 	Sequence      uint64
+	// instanceID is used to create a unique cache key for the org
+	instanceID string
 
 	Name   string
 	Domain string
@@ -110,7 +113,7 @@ func (q *Queries) OrgByID(ctx context.Context, shouldTriggerBulk bool, id string
 	ctx, span := tracing.NewSpan(ctx)
 	defer func() { span.EndWithError(err) }()
 
-	if org, ok := q.caches.org.Get(ctx, orgIndexByID, id); ok {
+	if org, ok := q.caches.org.Get(ctx, orgIndexByID, orgCacheKey(authz.GetInstance(ctx).InstanceID(), id)); ok {
 		return org, nil
 	}
 	defer func() {
@@ -147,6 +150,7 @@ func (q *Queries) OrgByID(ctx context.Context, shouldTriggerBulk bool, id string
 		ResourceOwner: foundOrg.Owner,
 		State:         domain_pkg.OrgState(foundOrg.State.State),
 		Sequence:      uint64(foundOrg.Sequence),
+		instanceID:    foundOrg.InstanceID,
 		Name:          foundOrg.Name,
 		Domain:        foundOrg.PrimaryDomain.Domain,
 	}, nil
@@ -183,7 +187,7 @@ func (q *Queries) OrgByPrimaryDomain(ctx context.Context, domain string) (org *O
 	ctx, span := tracing.NewSpan(ctx)
 	defer func() { span.EndWithError(err) }()
 
-	org, ok := q.caches.org.Get(ctx, orgIndexByPrimaryDomain, domain)
+	org, ok := q.caches.org.Get(ctx, orgIndexByPrimaryDomain, orgCacheKey(authz.GetInstance(ctx).InstanceID(), domain))
 	if ok {
 		return org, nil
 	}
@@ -416,6 +420,7 @@ func prepareOrgQuery() (sq.SelectBuilder, func(*sql.Row) (*Org, error)) {
 			OrgColumnResourceOwner.identifier(),
 			OrgColumnState.identifier(),
 			OrgColumnSequence.identifier(),
+			OrgColumnInstanceID.identifier(),
 			OrgColumnName.identifier(),
 			OrgColumnDomain.identifier(),
 		).
@@ -430,6 +435,7 @@ func prepareOrgQuery() (sq.SelectBuilder, func(*sql.Row) (*Org, error)) {
 				&o.ResourceOwner,
 				&o.State,
 				&o.Sequence,
+				&o.instanceID,
 				&o.Name,
 				&o.Domain,
 			)
@@ -507,15 +513,21 @@ const (
 func (o *Org) Keys(index orgIndex) []string {
 	switch index {
 	case orgIndexByID:
-		return []string{o.ID}
+		return []string{orgCacheKey(o.instanceID, o.ID)}
 	case orgIndexByPrimaryDomain:
-		return []string{o.Domain}
+		return []string{orgCacheKey(o.instanceID, o.Domain)}
 	case orgIndexUnspecified:
 	}
 	return nil
 }
 
+func orgCacheKey(instanceID, key string) string {
+	return instanceID + "-" + key
+}
+
 func (c *Caches) registerOrgInvalidation() {
-	invalidate := cacheInvalidationFunc(c.org, orgIndexByID, getAggregateID)
+	invalidate := cacheInvalidationFunc(c.org, orgIndexByID, func(aggregate *es.Aggregate) string {
+		return orgCacheKey(aggregate.InstanceID, aggregate.ID)
+	})
 	projection.OrgProjection.RegisterCacheInvalidation(invalidate)
 }

internal/query/org_test.go

@@ -50,6 +50,7 @@ var (
 		` projections.orgs1.resource_owner,` +
 		` projections.orgs1.org_state,` +
 		` projections.orgs1.sequence,` +
+		` projections.orgs1.instance_id,` +
 		` projections.orgs1.name,` +
 		` projections.orgs1.primary_domain` +
 		` FROM projections.orgs1`
@@ -60,6 +61,7 @@ var (
 		"resource_owner",
 		"org_state",
 		"sequence",
+		"instance_id",
 		"name",
 		"primary_domain",
 	}
@@ -242,6 +244,7 @@ func Test_OrgPrepares(t *testing.T) {
 						"ro",
 						domain.OrgStateActive,
 						uint64(20211108),
+						"instance-id",
 						"org-name",
 						"zitadel.ch",
 					},
@@ -254,6 +257,7 @@ func Test_OrgPrepares(t *testing.T) {
 				ResourceOwner: "ro",
 				State:         domain.OrgStateActive,
 				Sequence:      20211108,
+				instanceID:    "instance-id",
 				Name:          "org-name",
 				Domain:        "zitadel.ch",
 			},