[v18] feat: Add bot instance cache

lib/auth/accesspoint/accesspoint.go

@@ -74,6 +74,7 @@ type Config struct {
 	AccessMonitoringRules   services.AccessMonitoringRules
 	AppSession              services.AppSession
 	Apps                    services.Apps
+	BotInstance             services.BotInstance
 	ClusterConfig           services.ClusterConfiguration
 	CrownJewels             services.CrownJewels
 	DatabaseObjects         services.DatabaseObjects
@@ -194,6 +195,7 @@ func NewCache(cfg Config) (*cache.Cache, error) {
 		PluginStaticCredentials: cfg.PluginStaticCredentials,
 		GitServers:              cfg.GitServers,
 		HealthCheckConfig:       cfg.HealthCheckConfig,
+		BotInstanceService:      cfg.BotInstance,
 	}
 
 	return cache.New(cfg.Setup(cacheCfg))

lib/auth/authclient/api.go

@@ -1226,6 +1226,12 @@ type Cache interface {
 	// HealthCheckConfigReader defines methods for fetching health checkc config
 	// resources.
 	services.HealthCheckConfigReader
+
+	// GetBotInstance returns the specified BotInstance resource.
+	GetBotInstance(ctx context.Context, botName, instanceID string) (*machineidv1.BotInstance, error)
+
+	// ListBotInstances returns a page of BotInstance resources.
+	ListBotInstances(ctx context.Context, botName string, pageSize int, lastToken string, search string) ([]*machineidv1.BotInstance, string, error)
 }
 
 type NodeWrapper struct {

lib/auth/grpcserver.go

@@ -5368,6 +5368,7 @@ func NewGRPCServer(cfg GRPCServerConfig) (*GRPCServer, error) {
 
 	botInstanceService, err := machineidv1.NewBotInstanceService(machineidv1.BotInstanceServiceConfig{
 		Authorizer: cfg.Authorizer,
+		Cache:      cfg.AuthServer.Cache,
 		Backend:    cfg.AuthServer.Services.BotInstance,
 		Clock:      cfg.AuthServer.GetClock(),
 	})

lib/auth/helpers.go

@@ -552,6 +552,7 @@ func InitTestAuthCache(p TestAuthCacheParams) error {
 		PluginStaticCredentials: p.AuthServer.Services.PluginStaticCredentials,
 		GitServers:              p.AuthServer.Services.GitServers,
 		HealthCheckConfig:       p.AuthServer.Services.HealthCheckConfig,
+		BotInstance:             p.AuthServer.Services.BotInstance,
 	})
 	if err != nil {
 		return trace.Wrap(err)

lib/auth/machineid/machineidv1/bot_instance_service.go

@@ -48,10 +48,20 @@ const (
 	ExpiryMargin = time.Minute * 5
 )
 
+// BotInstancesCache is the subset of the cached resources that the Service queries.
+type BotInstancesCache interface {
+	// GetBotInstance returns the specified BotInstance resource.
+	GetBotInstance(ctx context.Context, botName, instanceID string) (*pb.BotInstance, error)
+
+	// ListBotInstances returns a page of BotInstance resources.
+	ListBotInstances(ctx context.Context, botName string, pageSize int, lastToken string, search string) ([]*pb.BotInstance, string, error)
+}
+
 // BotInstanceServiceConfig holds configuration options for the BotInstance gRPC
 // service.
 type BotInstanceServiceConfig struct {
 	Authorizer authz.Authorizer
+	Cache      BotInstancesCache
 	Backend    services.BotInstance
 	Logger     *slog.Logger
 	Clock      clockwork.Clock
@@ -64,6 +74,8 @@ func NewBotInstanceService(cfg BotInstanceServiceConfig) (*BotInstanceService, e
 		return nil, trace.BadParameter("backend service is required")
 	case cfg.Authorizer == nil:
 		return nil, trace.BadParameter("authorizer is required")
+	case cfg.Cache == nil:
+		return nil, trace.BadParameter("cache service is required")
 	}
 
 	if cfg.Logger == nil {
@@ -76,6 +88,7 @@ func NewBotInstanceService(cfg BotInstanceServiceConfig) (*BotInstanceService, e
 	return &BotInstanceService{
 		logger:     cfg.Logger,
 		authorizer: cfg.Authorizer,
+		cache:      cfg.Cache,
 		backend:    cfg.Backend,
 		clock:      cfg.Clock,
 	}, nil
@@ -87,6 +100,7 @@ type BotInstanceService struct {
 
 	backend    services.BotInstance
 	authorizer authz.Authorizer
+	cache      BotInstancesCache
 	logger     *slog.Logger
 	clock      clockwork.Clock
 }
@@ -124,7 +138,7 @@ func (b *BotInstanceService) GetBotInstance(ctx context.Context, req *pb.GetBotI
 		return nil, trace.Wrap(err)
 	}
 
-	res, err := b.backend.GetBotInstance(ctx, req.BotName, req.InstanceId)
+	res, err := b.cache.GetBotInstance(ctx, req.BotName, req.InstanceId)
 	if err != nil {
 		return nil, trace.Wrap(err)
 	}
@@ -143,7 +157,7 @@ func (b *BotInstanceService) ListBotInstances(ctx context.Context, req *pb.ListB
 		return nil, trace.Wrap(err)
 	}
 
-	res, nextToken, err := b.backend.ListBotInstances(ctx, req.FilterBotName, int(req.PageSize), req.PageToken, req.FilterSearchTerm)
+	res, nextToken, err := b.cache.ListBotInstances(ctx, req.FilterBotName, int(req.PageSize), req.PageToken, req.FilterSearchTerm)
 	if err != nil {
 		return nil, trace.Wrap(err)
 	}

lib/auth/machineid/machineidv1/bot_instance_service_test.go

@@ -334,6 +334,7 @@ func TestBotInstanceServiceSubmitHeartbeat(t *testing.T) {
 			backend := newBotInstanceBackend(t)
 			service, err := NewBotInstanceService(BotInstanceServiceConfig{
 				Backend: backend,
+				Cache:   backend,
 				Authorizer: authz.AuthorizerFunc(func(ctx context.Context) (*authz.Context, error) {
 					return &authz.Context{
 						Identity: identityGetterFn(func() tlsca.Identity {
@@ -391,6 +392,7 @@ func TestBotInstanceServiceSubmitHeartbeat_HeartbeatLimit(t *testing.T) {
 	backend := newBotInstanceBackend(t)
 	service, err := NewBotInstanceService(BotInstanceServiceConfig{
 		Backend: backend,
+		Cache:   backend,
 		Authorizer: authz.AuthorizerFunc(func(ctx context.Context) (*authz.Context, error) {
 			return &authz.Context{
 				Identity: identityGetterFn(func() tlsca.Identity {
@@ -583,6 +585,7 @@ func newBotInstanceService(
 	service, err := NewBotInstanceService(BotInstanceServiceConfig{
 		Authorizer: authorizer,
 		Backend:    backendService,
+		Cache:      backendService,
 	})
 	require.NoError(t, err)
 

lib/cache/bot_instance.go

@@ -0,0 +1,155 @@
+// Teleport
+// Copyright (C) 2025 Gravitational, Inc.
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+package cache
+
+import (
+	"context"
+	"slices"
+	"strings"
+
+	"github.com/gravitational/trace"
+	"google.golang.org/protobuf/proto"
+
+	"github.com/gravitational/teleport/api/defaults"
+	machineidv1 "github.com/gravitational/teleport/api/gen/proto/go/teleport/machineid/v1"
+	"github.com/gravitational/teleport/api/types"
+	"github.com/gravitational/teleport/api/utils/clientutils"
+	"github.com/gravitational/teleport/lib/services"
+)
+
+type botInstanceIndex string
+
+const (
+	botInstanceNameIndex botInstanceIndex = "name"
+)
+
+func keyForNameIndex(botInstance *machineidv1.BotInstance) string {
+	return makeNameIndexKey(
+		botInstance.GetSpec().GetBotName(),
+		botInstance.GetMetadata().GetName(),
+	)
+}
+
+func makeNameIndexKey(botName string, instanceID string) string {
+	return botName + "/" + instanceID
+}
+
+func newBotInstanceCollection(upstream services.BotInstance, w types.WatchKind) (*collection[*machineidv1.BotInstance, botInstanceIndex], error) {
+	if upstream == nil {
+		return nil, trace.BadParameter("missing parameter upstream (BotInstance)")
+	}
+
+	return &collection[*machineidv1.BotInstance, botInstanceIndex]{
+		store: newStore(
+			types.KindBotInstance,
+			proto.CloneOf[*machineidv1.BotInstance],
+			map[botInstanceIndex]func(*machineidv1.BotInstance) string{
+				// Index on a combination of bot name and instance name
+				botInstanceNameIndex: keyForNameIndex,
+			}),
+		fetcher: func(ctx context.Context, loadSecrets bool) ([]*machineidv1.BotInstance, error) {
+			var out []*machineidv1.BotInstance
+			clientutils.IterateResources(ctx,
+				func(ctx context.Context, limit int, start string) ([]*machineidv1.BotInstance, string, error) {
+					return upstream.ListBotInstances(ctx, "", limit, start, "")
+				},
+				func(hcc *machineidv1.BotInstance) error {
+					out = append(out, hcc)
+					return nil
+				},
+			)
+			return out, nil
+		},
+		watch: w,
+	}, nil
+}
+
+// GetBotInstance returns the specified BotInstance resource.
+func (c *Cache) GetBotInstance(ctx context.Context, botName, instanceID string) (*machineidv1.BotInstance, error) {
+	ctx, span := c.Tracer.Start(ctx, "cache/GetBotInstance")
+	defer span.End()
+
+	getter := genericGetter[*machineidv1.BotInstance, botInstanceIndex]{
+		cache:      c,
+		collection: c.collections.botInstances,
+		index:      botInstanceNameIndex,
+		upstreamGet: func(ctx context.Context, _ string) (*machineidv1.BotInstance, error) {
+			return c.Config.BotInstanceService.GetBotInstance(ctx, botName, instanceID)
+		},
+	}
+
+	out, err := getter.get(ctx, makeNameIndexKey(botName, instanceID))
+	return out, trace.Wrap(err)
+}
+
+// ListBotInstances returns a page of BotInstance resources.
+func (c *Cache) ListBotInstances(ctx context.Context, botName string, pageSize int, lastToken string, search string) ([]*machineidv1.BotInstance, string, error) {
+	ctx, span := c.Tracer.Start(ctx, "cache/ListBotInstances")
+	defer span.End()
+
+	lister := genericLister[*machineidv1.BotInstance, botInstanceIndex]{
+		cache:           c,
+		collection:      c.collections.botInstances,
+		index:           botInstanceNameIndex,
+		defaultPageSize: defaults.DefaultChunkSize,
+		upstreamList: func(ctx context.Context, limit int, start string) ([]*machineidv1.BotInstance, string, error) {
+			return c.Config.BotInstanceService.ListBotInstances(ctx, botName, limit, start, search)
+		},
+		filter: func(b *machineidv1.BotInstance) bool {
+			return matchBotInstance(b, botName, search)
+		},
+		nextToken: func(b *machineidv1.BotInstance) string {
+			return keyForNameIndex(b)
+		},
+	}
+	out, next, err := lister.list(ctx,
+		pageSize,
+		lastToken,
+	)
+	return out, next, trace.Wrap(err)
+}
+
+func matchBotInstance(b *machineidv1.BotInstance, botName string, search string) bool {
+	// If updating this, ensure it's consistent with the upstream search logic in `lib/services/local/bot_instance.go`.
+
+	if botName != "" && b.Spec.BotName != botName {
+		return false
+	}
+
+	if search == "" {
+		return true
+	}
+
+	latestHeartbeats := b.GetStatus().GetLatestHeartbeats()
+	heartbeat := b.Status.InitialHeartbeat // Use initial heartbeat as a fallback
+	if len(latestHeartbeats) > 0 {
+		heartbeat = latestHeartbeats[len(latestHeartbeats)-1]
+	}
+
+	values := []string{
+		b.Spec.BotName,
+		b.Spec.InstanceId,
+	}
+
+	if heartbeat != nil {
+		values = append(values, heartbeat.Hostname, heartbeat.JoinMethod, heartbeat.Version, "v"+heartbeat.Version)
+	}
+
+	return slices.ContainsFunc(values, func(val string) bool {
+		return strings.Contains(strings.ToLower(val), strings.ToLower(search))
+	})
+}