cmd/openolt-adapter/main.go

/*
* Copyright 2018-2024 Open Networking Foundation (ONF) and the ONF Contributors

* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at

* http://www.apache.org/licenses/LICENSE-2.0

* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
 */

// Package main invokes the application
package main

import (
	"context"
	"errors"
	"fmt"
	"net/http"
	"os"
	"os/signal"
	"syscall"
	"time"

	grpc_retry "github.com/grpc-ecosystem/go-grpc-middleware/retry"
	"github.com/prometheus/client_golang/prometheus/promhttp"
	codes "google.golang.org/grpc/codes"

	conf "github.com/opencord/voltha-lib-go/v7/pkg/config"
	"github.com/opencord/voltha-lib-go/v7/pkg/db/kvstore"
	"github.com/opencord/voltha-lib-go/v7/pkg/events"
	"github.com/opencord/voltha-lib-go/v7/pkg/events/eventif"
	vgrpc "github.com/opencord/voltha-lib-go/v7/pkg/grpc"
	"github.com/opencord/voltha-lib-go/v7/pkg/kafka"
	"github.com/opencord/voltha-lib-go/v7/pkg/log"
	"github.com/opencord/voltha-lib-go/v7/pkg/probe"
	"github.com/opencord/voltha-lib-go/v7/pkg/version"
	"github.com/opencord/voltha-openolt-adapter/internal/pkg/config"
	ac "github.com/opencord/voltha-openolt-adapter/internal/pkg/core"
	"github.com/opencord/voltha-protos/v5/go/adapter_service"
	ca "github.com/opencord/voltha-protos/v5/go/core_adapter"
	"github.com/opencord/voltha-protos/v5/go/core_service"
	"github.com/opencord/voltha-protos/v5/go/olt_inter_adapter_service"
	"github.com/opencord/voltha-protos/v5/go/voltha"
	"google.golang.org/grpc"
)

const (
	clusterMessagingService = "cluster-message-service"
	oltAdapterService       = "olt-adapter-service"
	kvService               = "kv-service"
	coreService             = "core-service"
	etcdStoreName           = "etcd"
)

type adapter struct {
	kafkaClient     kafka.Client
	kvClient        kvstore.Client
	eventProxy      eventif.EventProxy
	config          *config.AdapterFlags
	grpcServer      *vgrpc.GrpcServer
	oltAdapter      *ac.OpenOLT
	oltInterAdapter *ac.OpenOLTInterAdapter
	coreClient      *vgrpc.Client
	exitChannel     chan int
	instanceID      string
	halted          bool
}

func newAdapter(cf *config.AdapterFlags) *adapter {
	var a adapter
	a.instanceID = cf.InstanceID
	a.config = cf
	a.halted = false
	a.exitChannel = make(chan int, 1)
	return &a
}

func (a *adapter) start(ctx context.Context) {
	logger.Info(ctx, "Starting Core Adapter components")
	var err error

	var p *probe.Probe
	if value := ctx.Value(probe.ProbeContextKey); value != nil {
		if _, ok := value.(*probe.Probe); ok {
			p = value.(*probe.Probe)
			p.RegisterService(
				ctx,
				clusterMessagingService,
				kvService,
				oltAdapterService,
				coreService,
			)
		}
	}

	// Setup KV Client
	logger.Debugw(ctx, "create-kv-client", log.Fields{"kvstore": a.config.KVStoreType})
	if err = a.setKVClient(ctx); err != nil {
		logger.Fatalw(ctx, "error-setting-kv-client", log.Fields{"error": err})
	}

	if p != nil {
		p.UpdateStatus(ctx, kvService, probe.ServiceStatusRunning)
	}

	// Setup Log Config
	cm := conf.NewConfigManager(ctx, a.kvClient, a.config.KVStoreType, a.config.KVStoreAddress, a.config.KVStoreTimeout)

	go conf.StartLogLevelConfigProcessing(cm, ctx)
	go conf.StartLogFeaturesConfigProcessing(cm, ctx)

	// Setup Kafka Client
	if a.kafkaClient, err = newKafkaClient(ctx, "sarama", a.config); err != nil {
		logger.Fatalw(ctx, "Unsupported-common-client", log.Fields{"error": err})
	}

	// Start kafka communication with the broker
	if err = kafka.StartAndWaitUntilKafkaConnectionIsUp(ctx, a.kafkaClient, a.config.HeartbeatCheckInterval, clusterMessagingService); err != nil {
		logger.Fatal(ctx, "unable-to-connect-to-kafka")
	}

	// Create the event proxy to post events to KAFKA
	a.eventProxy = events.NewEventProxy(events.MsgClient(a.kafkaClient), events.MsgTopic(kafka.Topic{Name: a.config.EventTopic}))
	go func() {
		if err = a.eventProxy.Start(); err != nil {
			logger.Fatalw(ctx, "event-proxy-cannot-start", log.Fields{"error": err})
		}
	}()

	// Create the Core client to handle requests to the Core.  Note that the coreClient is an interface and needs to be
	// cast to the appropriate grpc client by invoking GetCoreGrpcClient on the a.coreClient
	if a.coreClient, err = vgrpc.NewClient(
		a.config.AdapterEndpoint,
		a.config.CoreEndpoint,
		"core_service.CoreService",
		a.coreRestarted); err != nil {
		logger.Fatal(ctx, "grpc-client-not-created")
	}
	// Start the core grpc client
	retryCodes := []codes.Code{
		codes.Unavailable,      // server is currently unavailable
		codes.DeadlineExceeded, // deadline for the operation was exceeded
	}
	// the backoff function sets the wait time bw each grpc retries, if not set it will take the deafault value of 50ms which is too low, the jitter sets the rpc retry wait time to be in a range of[PerRPCRetryTimeout-0.2, PerRPCRetryTimeout+0.2]
	backoffCtxOption := grpc_retry.WithBackoff(grpc_retry.BackoffLinearWithJitter(a.config.PerRPCRetryTimeout, 0.2))
	grpcRetryOptions := grpc_retry.UnaryClientInterceptor(grpc_retry.WithMax(a.config.MaxRetries), grpc_retry.WithPerRetryTimeout(a.config.PerRPCRetryTimeout), grpc_retry.WithCodes(retryCodes...), backoffCtxOption)
	logger.Debug(ctx, "Configuration values", log.Fields{"RETRY": a.config.MaxRetries, "TIMEOUT": a.config.PerRPCRetryTimeout})
	go a.coreClient.Start(ctx, getCoreServiceClientHandler, grpcRetryOptions)

	// Create the open OLT adapter
	if a.oltAdapter, err = a.startOpenOLT(ctx, a.coreClient, a.eventProxy, a.config, cm); err != nil {
		logger.Fatalw(ctx, "error-starting-openolt", log.Fields{"error": err})
	}

	// Create the open OLT Inter adapter adapter
	if a.oltInterAdapter, err = a.startOpenOLTInterAdapter(ctx, a.oltAdapter); err != nil {
		logger.Fatalw(ctx, "error-starting-openolt-inter-adapter", log.Fields{"error": err})
	}

	// Create and start the grpc server
	a.grpcServer = vgrpc.NewGrpcServer(a.config.GrpcAddress, nil, false, p)

	// Register the  adapter  service
	a.addAdapterService(ctx, a.grpcServer, a.oltAdapter)

	// Register the olt inter-adapter  service
	a.addOltInterAdapterService(ctx, a.grpcServer, a.oltInterAdapter)

	// Start the grpc server
	go a.startGRPCService(ctx, a.grpcServer, oltAdapterService)

	// Register this adapter to the Core - retries indefinitely
	if err = a.registerWithCore(ctx, coreService, -1); err != nil {
		logger.Fatal(ctx, "error-registering-with-core")
	}

	// check the readiness and liveliness and update the probe status
	a.checkServicesReadiness(ctx)
}

// TODO:  Any action the adapter needs to do following a Core restart?
func (a *adapter) coreRestarted(ctx context.Context, endPoint string) error {
	logger.Errorw(ctx, "core-restarted", log.Fields{"endpoint": endPoint})
	return nil
}

// getCoreServiceClientHandler is used to test whether the remote gRPC service is up
func getCoreServiceClientHandler(ctx context.Context, conn *grpc.ClientConn) interface{} {
	if conn == nil {
		return nil
	}
	return core_service.NewCoreServiceClient(conn)
}

/*
*
This function checks the liveliness and readiness of the kakfa and kv-client services
and update the status in the probe.
*/
func (a *adapter) checkServicesReadiness(ctx context.Context) {
	// checks the kafka readiness
	go kafka.MonitorKafkaReadiness(ctx, a.kafkaClient, a.config.LiveProbeInterval, a.config.NotLiveProbeInterval, clusterMessagingService)

	// checks the kv-store readiness
	go a.checkKvStoreReadiness(ctx)
}

/*
*
This function checks the liveliness and readiness of the kv-store service
and update the status in the probe.
*/
func (a *adapter) checkKvStoreReadiness(ctx context.Context) {
	// dividing the live probe interval by 2 to get updated status every 30s
	timeout := a.config.LiveProbeInterval / 2
	kvStoreChannel := make(chan bool, 1)

	timeoutCtx, cancelFunc := context.WithTimeout(ctx, 2*time.Second)
	kvStoreChannel <- a.kvClient.IsConnectionUp(timeoutCtx)
	cancelFunc()

	for {
		timeoutTimer := time.NewTimer(timeout)
		select {
		case liveliness := <-kvStoreChannel:
			if !liveliness {
				// kv-store not reachable or down, updating the status to not ready state
				probe.UpdateStatusFromContext(ctx, kvService, probe.ServiceStatusNotReady)
				timeout = a.config.NotLiveProbeInterval
			} else {
				// kv-store is reachable , updating the status to running state
				probe.UpdateStatusFromContext(ctx, kvService, probe.ServiceStatusRunning)
				timeout = a.config.LiveProbeInterval / 2
			}

			// Check if the timer has expired or not
			if !timeoutTimer.Stop() {
				<-timeoutTimer.C
			}

		case <-timeoutTimer.C:
			// Check the status of the kv-store. Use timeout of 2 seconds to avoid forever blocking
			timeoutCtx, cancelFunc := context.WithTimeout(ctx, 2*time.Second)

			kvStoreChannel <- a.kvClient.IsConnectionUp(timeoutCtx)
			// Cleanup cancel func resources
			cancelFunc()
		}
	}
}

func (a *adapter) stop(ctx context.Context) {
	// Stop leadership tracking
	a.halted = true

	// send exit signal
	a.exitChannel <- 0

	// Stop all grpc processing
	if err := a.oltAdapter.Stop(ctx); err != nil {
		logger.Errorw(ctx, "failure-stopping-olt-adapter-service", log.Fields{"error": err, "adapter": a.config.AdapterName})
	}
	if err := a.oltInterAdapter.Stop(ctx); err != nil {
		logger.Errorw(ctx, "failure-stopping-olt-inter-adapter-service", log.Fields{"error": err, "adapter": a.config.AdapterName})
	}

	// Cleanup - applies only if we had a kvClient
	if a.kvClient != nil {
		// Release all reservations
		if err := a.kvClient.ReleaseAllReservations(ctx); err != nil {
			logger.Infow(ctx, "fail-to-release-all-reservations", log.Fields{"error": err})
		}
		// Close the DB connection
		go a.kvClient.Close(ctx)
	}

	if a.eventProxy != nil {
		a.eventProxy.Stop()
	}

	if a.kafkaClient != nil {
		a.kafkaClient.Stop(ctx)
	}

	// Stop core client
	if a.coreClient != nil {
		a.coreClient.Stop(ctx)
	}

	logger.Info(ctx, "main-stop-processing-complete")

	// TODO: Stop child devices connections

	// TODO:  More cleanup
}

func newKVClient(ctx context.Context, storeType, address string, timeout time.Duration) (kvstore.Client, error) {
	logger.Infow(ctx, "kv-store-type", log.Fields{"store": storeType})
	switch storeType {
	case etcdStoreName:
		return kvstore.NewEtcdClient(ctx, address, timeout, log.FatalLevel)
	case "redis":
		return kvstore.NewRedisClient(address, timeout, false)
	case "redis-sentinel":
		return kvstore.NewRedisClient(address, timeout, true)
	}
	return nil, errors.New("unsupported-kv-store")
}

func newKafkaClient(ctx context.Context, clientType string, config *config.AdapterFlags) (kafka.Client, error) {
	logger.Infow(ctx, "common-client-type", log.Fields{"client": clientType})
	switch clientType {
	case "sarama":
		return kafka.NewSaramaClient(
			kafka.Address(config.KafkaClusterAddress),
			kafka.ProducerReturnOnErrors(true),
			kafka.ProducerReturnOnSuccess(true),
			kafka.ProducerMaxRetries(config.ProducerRetryMax),
			kafka.ProducerRetryBackoff(time.Millisecond*30),
			kafka.MetadatMaxRetries(config.MetadataRetryMax)), nil
	}

	return nil, errors.New("unsupported-client-type")
}

func (a *adapter) setKVClient(ctx context.Context) error {
	client, err := newKVClient(ctx, a.config.KVStoreType, a.config.KVStoreAddress, a.config.KVStoreTimeout)
	if err != nil {
		a.kvClient = nil
		return err
	}
	a.kvClient = client

	return nil
}

// startGRPCService creates the grpc service handlers, registers it to the grpc server and starts the server
func (a *adapter) startGRPCService(ctx context.Context, server *vgrpc.GrpcServer, serviceName string) {
	logger.Infow(ctx, "starting-grpc-service", log.Fields{"service": serviceName})

	probe.UpdateStatusFromContext(ctx, serviceName, probe.ServiceStatusRunning)
	logger.Infow(ctx, "grpc-service-started", log.Fields{"service": serviceName})

	server.Start(ctx)
	probe.UpdateStatusFromContext(ctx, serviceName, probe.ServiceStatusStopped)
}

func (a *adapter) addAdapterService(ctx context.Context, server *vgrpc.GrpcServer, handler adapter_service.AdapterServiceServer) {
	logger.Info(ctx, "adding-adapter-service")

	server.AddService(func(gs *grpc.Server) {
		adapter_service.RegisterAdapterServiceServer(gs, handler)
	})
}

func (a *adapter) addOltInterAdapterService(ctx context.Context, server *vgrpc.GrpcServer, handler olt_inter_adapter_service.OltInterAdapterServiceServer) {
	logger.Info(ctx, "adding-olt-inter-adapter-service")

	server.AddService(func(gs *grpc.Server) {
		olt_inter_adapter_service.RegisterOltInterAdapterServiceServer(gs, handler)
	})
}

func (a *adapter) startOpenOLT(ctx context.Context, cc *vgrpc.Client, ep eventif.EventProxy,
	cfg *config.AdapterFlags, cm *conf.ConfigManager) (*ac.OpenOLT, error) {
	logger.Info(ctx, "starting-open-olt")
	var err error
	sOLT := ac.NewOpenOLT(ctx, cc, ep, cfg, cm)

	if err = sOLT.Start(ctx); err != nil {
		return nil, err
	}

	logger.Info(ctx, "open-olt-started")
	return sOLT, nil
}

func (a *adapter) startOpenOLTInterAdapter(ctx context.Context, oo *ac.OpenOLT) (*ac.OpenOLTInterAdapter, error) {
	logger.Info(ctx, "starting-open-olt-inter-adapter")
	var err error
	sOLTInterAdapter := ac.NewOpenOLTInterAdapter(oo)

	if err = sOLTInterAdapter.Start(ctx); err != nil {
		return nil, err
	}

	logger.Info(ctx, "open-olt-inter-adapter-started")
	return sOLTInterAdapter, nil
}

func (a *adapter) registerWithCore(ctx context.Context, serviceName string, retries int) error {
	adapterID := fmt.Sprintf("openolt_%d", a.config.CurrentReplica)
	logger.Infow(ctx, "registering-with-core", log.Fields{
		"adapterID":      adapterID,
		"currentReplica": a.config.CurrentReplica,
		"totalReplicas":  a.config.TotalReplicas,
	})
	adapterDescription := &voltha.Adapter{
		Id:      adapterID, // Unique name for the device type
		Vendor:  "VOLTHA OpenOLT",
		Version: version.VersionInfo.Version,
		// The Endpoint refers to the address this service is listening on.
		Endpoint:       a.config.AdapterEndpoint,
		Type:           "openolt",
		CurrentReplica: int32(a.config.CurrentReplica),
		TotalReplicas:  int32(a.config.TotalReplicas),
	}
	types := []*voltha.DeviceType{{
		Id:                          "openolt",
		AdapterType:                 "openolt", // Type of the adapter that handles device type
		Adapter:                     "openolt", // Deprecated attribute
		AcceptsBulkFlowUpdate:       false,     // Currently openolt adapter does not support bulk flow handling
		AcceptsAddRemoveFlowUpdates: true}}
	deviceTypes := &voltha.DeviceTypes{Items: types}
	count := 0
	for {
		gClient, err := a.coreClient.GetCoreServiceClient()
		if gClient != nil {
			if _, err = gClient.RegisterAdapter(log.WithSpanFromContext(context.TODO(), ctx), &ca.AdapterRegistration{
				Adapter: adapterDescription,
				DTypes:  deviceTypes}); err == nil {
				break
			}
		}
		logger.Warnw(ctx, "registering-with-core-failed", log.Fields{"endpoint": a.config.CoreEndpoint, "error": err, "count": count, "gclient": gClient})
		if retries == count {
			return err
		}
		count++
		// Take a nap before retrying
		time.Sleep(2 * time.Second)
	}
	probe.UpdateStatusFromContext(ctx, serviceName, probe.ServiceStatusRunning)
	logger.Info(ctx, "registered-with-core")
	return nil
}

func waitForExit(ctx context.Context) int {
	signalChannel := make(chan os.Signal, 1)
	signal.Notify(signalChannel,
		syscall.SIGHUP,
		syscall.SIGINT,
		syscall.SIGTERM,
		syscall.SIGQUIT)

	exitChannel := make(chan int)

	go func() {
		s := <-signalChannel
		switch s {
		case syscall.SIGHUP,
			syscall.SIGINT,
			syscall.SIGTERM,
			syscall.SIGQUIT:
			logger.Infow(ctx, "closing-signal-received", log.Fields{"signal": s})
			exitChannel <- 0
		default:
			logger.Infow(ctx, "unexpected-signal-received", log.Fields{"signal": s})
			exitChannel <- 1
		}
	}()

	code := <-exitChannel
	return code
}

func printBanner() {
	fmt.Println(`   ____                     ____  _   _______ `)
	fmt.Println(`  / _  \                   / __ \| | |__   __|`)
	fmt.Println(` | |  | |_ __   ___ _ __  | |  | | |    | |   `)
	fmt.Println(` | |  | | '_ \ / _ \ '_ \ | |  | | |    | |   `)
	fmt.Println(` | |__| | |_) |  __/ | | || |__| | |____| |   `)
	fmt.Println(`  \____/| .__/ \___|_| |_| \____/|______|_|   `)
	fmt.Println(`        | |                                   `)
	fmt.Println(`        |_|                                   `)
	fmt.Println(`                                              `)
}

func printVersion() {
	fmt.Println("VOLTHA OpenOLT Adapter")
	fmt.Println(version.VersionInfo.String("  "))
}

func main() {
	ctx := context.Background()
	start := time.Now()

	cf := config.NewAdapterFlags()
	cf.ParseCommandArguments()

	// Setup logging

	logLevel, err := log.StringToLogLevel(cf.LogLevel)
	if err != nil {
		logger.Fatalf(ctx, "Cannot setup logging, %s", err)
	}

	// Setup default logger - applies for packages that do not have specific logger set
	if _, err = log.SetDefaultLogger(log.JSON, logLevel, log.Fields{"instanceId": cf.InstanceID}); err != nil {
		logger.With(log.Fields{"error": err}).Fatal(ctx, "Cannot setup logging")
	}

	// Update all loggers (provisionned via init) with a common field
	if err = log.UpdateAllLoggers(log.Fields{"instanceId": cf.InstanceID}); err != nil {
		logger.With(log.Fields{"error": err}).Fatal(ctx, "Cannot setup logging")
	}

	log.SetAllLogLevel(logLevel)

	realMain()

	defer func() {
		err = log.CleanUp()
		if err != nil {
			logger.Errorw(context.Background(), "unable-to-flush-any-buffered-log-entries", log.Fields{"error": err})
		}
	}()

	// Print version / build information and exit
	if cf.DisplayVersionOnly {
		printVersion()
		return
	}

	// Print banner if specified
	if cf.Banner {
		printBanner()
	}

	logger.Infow(ctx, "config", log.Fields{"config": *cf})

	ctx, cancel := context.WithCancel(context.Background())
	defer cancel()

	ad := newAdapter(cf)
	http.Handle("/metrics", promhttp.Handler())
	go func() {
		logger.Infof(ctx, "Metrics available at %s/metrics", ad.config.PrometheusAddress)
		// Create HTTP server with explicit timeouts to prevent slowloris attacks and resource exhaustion.
		// Using http.ListenAndServe() directly doesn't allow setting timeouts, which is a security risk.
		// The server uses http.DefaultServeMux (nil handler) which includes the /metrics endpoint registered above.
		metricsServer := &http.Server{
			Addr:              ad.config.PrometheusAddress,
			Handler:           nil,
			ReadHeaderTimeout: 10 * time.Second,
			ReadTimeout:       30 * time.Second,
			WriteTimeout:      30 * time.Second,
			IdleTimeout:       120 * time.Second,
		}
		if err := metricsServer.ListenAndServe(); err != nil {
			logger.Errorw(ctx, "failed to start metrics HTTP server: ", log.Fields{"error": err})
		}
	}()

	p := &probe.Probe{}
	go p.ListenAndServe(ctx, ad.config.ProbeAddress)

	probeCtx := context.WithValue(ctx, probe.ProbeContextKey, p)

	closer, err := log.GetGlobalLFM().InitTracingAndLogCorrelation(cf.TraceEnabled, cf.TraceAgentAddress, cf.LogCorrelationEnabled)
	if err != nil {
		logger.Warnw(ctx, "unable-to-initialize-tracing-and-log-correlation-module", log.Fields{"error": err})
	} else {
		defer log.TerminateTracing(closer)
	}

	go ad.start(probeCtx)

	code := waitForExit(ctx)
	logger.Infow(ctx, "received-a-closing-signal", log.Fields{"code": code})

	// Use context with cancel as etcd-client stop could take more time sometimes to stop slowing down container shutdown.
	ctxWithCancel, cancelFunc := context.WithCancel(ctx)
	// Cleanup before leaving
	ad.stop(ctxWithCancel)
	// Will halt any long-running stop routine gracefully
	cancelFunc()

	elapsed := time.Since(start)
	logger.Infow(ctx, "run-time", log.Fields{"instanceId": ad.config.InstanceID, "time": elapsed / time.Second})
}

// [EOF]