diff --git a/integration/README.rst b/integration/README.rst
index aebbd57ef55..e9e04582c83 100644
--- a/integration/README.rst
+++ b/integration/README.rst
@@ -44,8 +44,7 @@ your machine.
 Caveats, Gotchas, and Good-To-Knows
 -----------------------------------
 * The tests in this repository only exercise components that originate in the Fabric repository.
-* Currently, docker is only used in the ginkgo tests when using chaincode, kafka, zookeeper,
-  or couchdb images.
+* Currently, docker is only used in the ginkgo tests when using chaincode or couchdb images.
 
 
 Getting Started
diff --git a/integration/ledger/reset_rollback_test.go b/integration/ledger/reset_rollback_test.go
index d44b967e0ef..df9d9182cd8 100644
--- a/integration/ledger/reset_rollback_test.go
+++ b/integration/ledger/reset_rollback_test.go
@@ -137,7 +137,7 @@ var _ = Describe("rollback, reset, pause, resume, and unjoin peer node commands"
 		helper.assertPausedChannel(org3peer0)
 
 		By("Checking preResetHeightFile exists for a paused channel that is also rolled back or reset")
-		setup.startBrokerAndOrderer()
+		setup.startOrderer()
 		preResetHeightFile := filepath.Join(setup.network.PeerLedgerDir(org3peer0), "chains/chains", helper.channelID, "__preResetHeight")
 		Expect(preResetHeightFile).To(BeARegularFile())
 
@@ -153,7 +153,7 @@ var _ = Describe("rollback, reset, pause, resume, and unjoin peer node commands"
 		}
 
 		By("Bringing the peers to recent height by starting the orderer")
-		setup.startBrokerAndOrderer()
+		setup.startOrderer()
 		for _, peer := range setup.peers {
 			By("Verifying endorsement is enabled and preResetHeightFile is removed on peer " + peer.ID())
 			helper.waitUntilEndorserEnabled(peer)
@@ -263,7 +263,6 @@ type setup struct {
 	peerProcess    []ifrit.Process
 	orderer        *nwo.Orderer
 	ordererProcess ifrit.Process
-	brokerProcess  ifrit.Process
 }
 
 func initThreeOrgsSetup() *setup {
@@ -291,7 +290,7 @@ func initThreeOrgsSetup() *setup {
 		channelID: "testchannel",
 	}
 
-	setup.startBrokerAndOrderer()
+	setup.startOrderer()
 
 	setup.startPeer(peers[0])
 	setup.startPeer(peers[1])
@@ -319,10 +318,6 @@ func (s *setup) terminateAllProcess() {
 	Eventually(s.ordererProcess.Wait(), s.network.EventuallyTimeout).Should(Receive())
 	s.ordererProcess = nil
 
-	s.brokerProcess.Signal(syscall.SIGTERM)
-	Eventually(s.brokerProcess.Wait(), s.network.EventuallyTimeout).Should(Receive())
-	s.brokerProcess = nil
-
 	for _, p := range s.peerProcess {
 		p.Signal(syscall.SIGTERM)
 		Eventually(p.Wait(), s.network.EventuallyTimeout).Should(Receive())
@@ -351,12 +346,7 @@ func (s *setup) startPeer(peer *nwo.Peer) {
 	s.peerProcess = append(s.peerProcess, peerProcess)
 }
 
-func (s *setup) startBrokerAndOrderer() {
-	brokerRunner := s.network.BrokerGroupRunner()
-	brokerProcess := ifrit.Invoke(brokerRunner)
-	Eventually(brokerProcess.Ready(), s.network.EventuallyTimeout).Should(BeClosed())
-	s.brokerProcess = brokerProcess
-
+func (s *setup) startOrderer() {
 	ordererRunner := s.network.OrdererGroupRunner()
 	ordererProcess := ifrit.Invoke(ordererRunner)
 	Eventually(ordererProcess.Ready(), s.network.EventuallyTimeout).Should(BeClosed())
diff --git a/integration/ledger/snapshot_test.go b/integration/ledger/snapshot_test.go
index b412f0bfd73..385116538ed 100644
--- a/integration/ledger/snapshot_test.go
+++ b/integration/ledger/snapshot_test.go
@@ -636,8 +636,8 @@ func initAndStartFourOrgsNetwork() *setup {
 	}
 	Expect(setup.testDir).To(Equal(setup.network.RootDir))
 
-	By("starting broker and orderer")
-	setup.startBrokerAndOrderer()
+	By("starting orderer")
+	setup.startOrderer()
 
 	By("starting peers")
 	setup.startPeers()
diff --git a/integration/nwo/network.go b/integration/nwo/network.go
index 8036a6c2611..18dc0336f7c 100644
--- a/integration/nwo/network.go
+++ b/integration/nwo/network.go
@@ -18,7 +18,6 @@ import (
 	"path/filepath"
 	"runtime"
 	"sort"
-	"strconv"
 	"strings"
 	"sync"
 	"syscall"
@@ -65,13 +64,10 @@ type Consortium struct {
 	Organizations []string `yaml:"organizations,omitempty"`
 }
 
-// Consensus indicates the orderer types and how many broker and zookeeper
-// instances.
+// Consensus indicates the orderer types.
 type Consensus struct {
 	Type                        string `yaml:"type,omitempty"`
 	BootstrapMethod             string `yaml:"bootstrap_method,omitempty"`
-	Brokers                     int    `yaml:"brokers,omitempty"`
-	ZooKeepers                  int    `yaml:"zookeepers,omitempty"`
 	ChannelParticipationEnabled bool   `yaml:"channel_participation_enabled,omitempty"`
 }
 
@@ -157,7 +153,6 @@ type Network struct {
 	TLSEnabled            bool
 	GatewayEnabled        bool
 
-	PortsByBrokerID  map[string]Ports
 	PortsByOrdererID map[string]Ports
 	PortsByPeerID    map[string]Ports
 	Organizations    []*Organization
@@ -188,7 +183,6 @@ func New(c *Config, rootDir string, dockerClient *docker.Client, startPort int,
 		NetworkID:         runner.UniqueName(),
 		EventuallyTimeout: time.Minute,
 		MetricsProvider:   "prometheus",
-		PortsByBrokerID:   map[string]Ports{},
 		PortsByOrdererID:  map[string]Ports{},
 		PortsByPeerID:     map[string]Ports{},
 
@@ -230,14 +224,6 @@ func New(c *Config, rootDir string, dockerClient *docker.Client, startPort int,
 		network.SessionCreateInterval = time.Second
 	}
 
-	for i := 0; i < network.Consensus.Brokers; i++ {
-		ports := Ports{}
-		for _, portName := range BrokerPortNames() {
-			ports[portName] = network.ReservePort()
-		}
-		network.PortsByBrokerID[strconv.Itoa(i)] = ports
-	}
-
 	for _, o := range c.Orderers {
 		ports := Ports{}
 		for _, portName := range OrdererPortNames() {
@@ -1245,87 +1231,6 @@ func (n *Network) Osnadmin(command Command) (*gexec.Session, error) {
 	return n.StartSession(cmd, command.SessionName())
 }
 
-// ZooKeeperRunner returns a runner for a ZooKeeper instance.
-func (n *Network) ZooKeeperRunner(idx int) *runner.ZooKeeper {
-	colorCode := n.nextColor()
-	name := fmt.Sprintf("zookeeper-%d-%s", idx, n.NetworkID)
-
-	return &runner.ZooKeeper{
-		ZooMyID:     idx + 1, //  IDs must be between 1 and 255
-		Client:      n.DockerClient,
-		Name:        name,
-		NetworkName: n.NetworkID,
-		OutputStream: gexec.NewPrefixedWriter(
-			fmt.Sprintf("\x1b[32m[o]\x1b[%s[%s]\x1b[0m ", colorCode, name),
-			ginkgo.GinkgoWriter,
-		),
-		ErrorStream: gexec.NewPrefixedWriter(
-			fmt.Sprintf("\x1b[91m[e]\x1b[%s[%s]\x1b[0m ", colorCode, name),
-			ginkgo.GinkgoWriter,
-		),
-	}
-}
-
-func (n *Network) minBrokersInSync() int {
-	if n.Consensus.Brokers < 2 {
-		return n.Consensus.Brokers
-	}
-	return 2
-}
-
-func (n *Network) defaultBrokerReplication() int {
-	if n.Consensus.Brokers < 3 {
-		return n.Consensus.Brokers
-	}
-	return 3
-}
-
-// BrokerRunner returns a runner for a Kafka broker instance.
-func (n *Network) BrokerRunner(id int, zookeepers []string) *runner.Kafka {
-	colorCode := n.nextColor()
-	name := fmt.Sprintf("kafka-%d-%s", id, n.NetworkID)
-
-	return &runner.Kafka{
-		BrokerID:                 id + 1,
-		Client:                   n.DockerClient,
-		AdvertisedListeners:      "127.0.0.1",
-		HostPort:                 int(n.PortsByBrokerID[strconv.Itoa(id)][HostPort]),
-		Name:                     name,
-		NetworkName:              n.NetworkID,
-		MinInsyncReplicas:        n.minBrokersInSync(),
-		DefaultReplicationFactor: n.defaultBrokerReplication(),
-		ZooKeeperConnect:         strings.Join(zookeepers, ","),
-		OutputStream: gexec.NewPrefixedWriter(
-			fmt.Sprintf("\x1b[32m[o]\x1b[%s[%s]\x1b[0m ", colorCode, name),
-			ginkgo.GinkgoWriter,
-		),
-		ErrorStream: gexec.NewPrefixedWriter(
-			fmt.Sprintf("\x1b[91m[e]\x1b[%s[%s]\x1b[0m ", colorCode, name),
-			ginkgo.GinkgoWriter,
-		),
-	}
-}
-
-// BrokerGroupRunner returns a runner that manages the processes that make up
-// the Kafka broker network for fabric.
-func (n *Network) BrokerGroupRunner() ifrit.Runner {
-	members := grouper.Members{}
-	zookeepers := []string{}
-
-	for i := 0; i < n.Consensus.ZooKeepers; i++ {
-		zk := n.ZooKeeperRunner(i)
-		zookeepers = append(zookeepers, fmt.Sprintf("%s:2181", zk.Name))
-		members = append(members, grouper.Member{Name: zk.Name, Runner: zk})
-	}
-
-	for i := 0; i < n.Consensus.Brokers; i++ {
-		kafka := n.BrokerRunner(i, zookeepers)
-		members = append(members, grouper.Member{Name: kafka.Name, Runner: kafka})
-	}
-
-	return grouper.NewOrdered(syscall.SIGTERM, members)
-}
-
 // OrdererRunner returns an ifrit.Runner for the specified orderer. The runner
 // can be used to start and manage an orderer process.
 func (n *Network) OrdererRunner(o *Orderer, env ...string) *ginkgomon.Runner {
@@ -1342,12 +1247,6 @@ func (n *Network) OrdererRunner(o *Orderer, env ...string) *ginkgomon.Runner {
 		StartCheckTimeout: 15 * time.Second,
 	}
 
-	// After consensus-type migration, the #brokers is >0, but the type is etcdraft
-	if n.Consensus.Type == "kafka" && n.Consensus.Brokers != 0 {
-		config.StartCheck = "Start phase completed successfully"
-		config.StartCheckTimeout = 30 * time.Second
-	}
-
 	return ginkgomon.New(config)
 }
 
@@ -1396,7 +1295,6 @@ func (n *Network) PeerGroupRunner() ifrit.Runner {
 // entire fabric network.
 func (n *Network) NetworkGroupRunner() ifrit.Runner {
 	members := grouper.Members{
-		{Name: "brokers", Runner: n.BrokerGroupRunner()},
 		{Name: "orderers", Runner: n.OrdererGroupRunner()},
 		{Name: "peers", Runner: n.PeerGroupRunner()},
 	}
@@ -1786,21 +1684,6 @@ func OrdererPortNames() []PortName {
 	return []PortName{ListenPort, ProfilePort, OperationsPort, ClusterPort, AdminPort}
 }
 
-// BrokerPortNames returns the list of ports that need to be reserved for a
-// Kafka broker.
-func BrokerPortNames() []PortName {
-	return []PortName{HostPort}
-}
-
-// BrokerAddresses returns the list of broker addresses for the network.
-func (n *Network) BrokerAddresses(portName PortName) []string {
-	addresses := []string{}
-	for _, ports := range n.PortsByBrokerID {
-		addresses = append(addresses, fmt.Sprintf("127.0.0.1:%d", ports[portName]))
-	}
-	return addresses
-}
-
 // OrdererAddress returns the address (host and port) exposed by the Orderer
 // for the named port. Command line tools should use the returned address when
 // connecting to the orderer.
diff --git a/integration/nwo/runner/kafka.go b/integration/nwo/runner/kafka.go
deleted file mode 100644
index 91b0320064e..00000000000
--- a/integration/nwo/runner/kafka.go
+++ /dev/null
@@ -1,304 +0,0 @@
-/*
-Copyright IBM Corp. All Rights Reserved.
-
-SPDX-License-Identifier: Apache-2.0
-*/
-
-package runner
-
-import (
-	"context"
-	"fmt"
-	"io"
-	"net"
-	"os"
-	"strconv"
-	"sync"
-	"time"
-
-	docker "github.com/fsouza/go-dockerclient"
-	"github.com/pkg/errors"
-	"github.com/tedsuo/ifrit"
-)
-
-const KafkaDefaultImage = "confluentinc/cp-kafka:5.3.1"
-
-// Kafka manages the execution of an instance of a dockerized CouchDB
-// for tests.
-type Kafka struct {
-	Client        *docker.Client
-	Image         string
-	HostIP        string
-	HostPort      int
-	ContainerPort docker.Port
-	Name          string
-	NetworkName   string
-	StartTimeout  time.Duration
-
-	MessageMaxBytes              int
-	ReplicaFetchMaxBytes         int
-	UncleanLeaderElectionEnable  bool
-	DefaultReplicationFactor     int
-	MinInsyncReplicas            int
-	BrokerID                     int
-	ZooKeeperConnect             string
-	ReplicaFetchResponseMaxBytes int
-	AdvertisedListeners          string
-
-	ErrorStream  io.Writer
-	OutputStream io.Writer
-
-	ContainerID      string
-	HostAddress      string
-	ContainerAddress string
-	Address          string
-
-	mutex   sync.Mutex
-	stopped bool
-}
-
-// Run runs a Kafka container. It implements the ifrit.Runner interface
-func (k *Kafka) Run(sigCh <-chan os.Signal, ready chan<- struct{}) error {
-	if k.Image == "" {
-		k.Image = KafkaDefaultImage
-	}
-
-	if k.Name == "" {
-		k.Name = DefaultNamer()
-	}
-
-	if k.HostIP == "" {
-		k.HostIP = "127.0.0.1"
-	}
-
-	if k.ContainerPort == docker.Port("") {
-		k.ContainerPort = docker.Port("9092/tcp")
-	}
-
-	if k.StartTimeout == 0 {
-		k.StartTimeout = DefaultStartTimeout
-	}
-
-	if k.Client == nil {
-		client, err := docker.NewClientFromEnv()
-		if err != nil {
-			return err
-		}
-		k.Client = client
-	}
-
-	if k.DefaultReplicationFactor == 0 {
-		k.DefaultReplicationFactor = 1
-	}
-
-	if k.MinInsyncReplicas == 0 {
-		k.MinInsyncReplicas = 1
-	}
-
-	if k.ZooKeeperConnect == "" {
-		k.ZooKeeperConnect = "zookeeper:2181/kafka"
-	}
-
-	if k.MessageMaxBytes == 0 {
-		k.MessageMaxBytes = 1000012
-	}
-
-	if k.ReplicaFetchMaxBytes == 0 {
-		k.ReplicaFetchMaxBytes = 1048576
-	}
-
-	if k.ReplicaFetchResponseMaxBytes == 0 {
-		k.ReplicaFetchResponseMaxBytes = 10485760
-	}
-
-	containerOptions := docker.CreateContainerOptions{
-		Name: k.Name,
-		Config: &docker.Config{
-			Image: k.Image,
-			Env:   k.buildEnv(),
-		},
-		HostConfig: &docker.HostConfig{
-			AutoRemove: true,
-			PortBindings: map[docker.Port][]docker.PortBinding{
-				k.ContainerPort: {{
-					HostIP:   k.HostIP,
-					HostPort: strconv.Itoa(k.HostPort),
-				}},
-			},
-		},
-	}
-
-	if k.NetworkName != "" {
-		nw, err := k.Client.NetworkInfo(k.NetworkName)
-		if err != nil {
-			return err
-		}
-
-		containerOptions.NetworkingConfig = &docker.NetworkingConfig{
-			EndpointsConfig: map[string]*docker.EndpointConfig{
-				k.NetworkName: {
-					NetworkID: nw.ID,
-				},
-			},
-		}
-	}
-
-	container, err := k.Client.CreateContainer(containerOptions)
-	if err != nil {
-		return err
-	}
-	k.ContainerID = container.ID
-
-	err = k.Client.StartContainer(container.ID, nil)
-	if err != nil {
-		return err
-	}
-	defer k.Stop()
-
-	container, err = k.Client.InspectContainer(container.ID)
-	if err != nil {
-		return err
-	}
-
-	k.HostAddress = net.JoinHostPort(
-		container.NetworkSettings.Ports[k.ContainerPort][0].HostIP,
-		container.NetworkSettings.Ports[k.ContainerPort][0].HostPort,
-	)
-	k.ContainerAddress = net.JoinHostPort(
-		container.NetworkSettings.Networks[k.NetworkName].IPAddress,
-		k.ContainerPort.Port(),
-	)
-
-	logContext, cancelLogs := context.WithCancel(context.Background())
-	defer cancelLogs()
-	go k.streamLogs(logContext)
-
-	containerExit := k.wait()
-	ctx, cancel := context.WithTimeout(context.Background(), k.StartTimeout)
-	defer cancel()
-
-	select {
-	case <-ctx.Done():
-		return errors.Wrapf(ctx.Err(), "kafka broker in container %s did not start", k.ContainerID)
-	case <-containerExit:
-		return errors.New("container exited before ready")
-	case <-k.ready(ctx, k.ContainerAddress):
-		k.Address = k.ContainerAddress
-	case <-k.ready(ctx, k.HostAddress):
-		k.Address = k.HostAddress
-	}
-
-	cancel()
-	close(ready)
-
-	for {
-		select {
-		case err := <-containerExit:
-			return err
-		case <-sigCh:
-			if err := k.Stop(); err != nil {
-				return err
-			}
-		}
-	}
-}
-
-func (k *Kafka) buildEnv() []string {
-	env := []string{
-		"KAFKA_LOG_RETENTION_MS=-1",
-		//"KAFKA_AUTO_CREATE_TOPICS_ENABLE=false",
-		fmt.Sprintf("KAFKA_MESSAGE_MAX_BYTES=%d", k.MessageMaxBytes),
-		fmt.Sprintf("KAFKA_REPLICA_FETCH_MAX_BYTES=%d", k.ReplicaFetchMaxBytes),
-		fmt.Sprintf("KAFKA_UNCLEAN_LEADER_ELECTION_ENABLE=%s", strconv.FormatBool(k.UncleanLeaderElectionEnable)),
-		fmt.Sprintf("KAFKA_DEFAULT_REPLICATION_FACTOR=%d", k.DefaultReplicationFactor),
-		fmt.Sprintf("KAFKA_MIN_INSYNC_REPLICAS=%d", k.MinInsyncReplicas),
-		fmt.Sprintf("KAFKA_BROKER_ID=%d", k.BrokerID),
-		fmt.Sprintf("KAFKA_ZOOKEEPER_CONNECT=%s", k.ZooKeeperConnect),
-		fmt.Sprintf("KAFKA_REPLICA_FETCH_RESPONSE_MAX_BYTES=%d", k.ReplicaFetchResponseMaxBytes),
-		fmt.Sprintf("KAFKA_ADVERTISED_LISTENERS=EXTERNAL://localhost:%d,%s://%s:9093", k.HostPort, k.NetworkName, k.Name),
-		fmt.Sprintf("KAFKA_LISTENERS=EXTERNAL://0.0.0.0:9092,%s://0.0.0.0:9093", k.NetworkName),
-		fmt.Sprintf("KAFKA_LISTENER_SECURITY_PROTOCOL_MAP=EXTERNAL:PLAINTEXT,%s:PLAINTEXT", k.NetworkName),
-		fmt.Sprintf("KAFKA_INTER_BROKER_LISTENER_NAME=%s", k.NetworkName),
-	}
-	return env
-}
-
-func (k *Kafka) ready(ctx context.Context, addr string) <-chan struct{} {
-	readyCh := make(chan struct{})
-	go func() {
-		ticker := time.NewTicker(100 * time.Millisecond)
-		defer ticker.Stop()
-
-		for {
-			conn, err := net.DialTimeout("tcp", addr, 50*time.Millisecond)
-			if err == nil {
-				conn.Close()
-				close(readyCh)
-				return
-			}
-
-			select {
-			case <-ticker.C:
-			case <-ctx.Done():
-				return
-			}
-		}
-	}()
-
-	return readyCh
-}
-
-func (k *Kafka) wait() <-chan error {
-	exitCh := make(chan error)
-	go func() {
-		exitCode, err := k.Client.WaitContainer(k.ContainerID)
-		if err == nil {
-			err = fmt.Errorf("kafka: process exited with %d", exitCode)
-		}
-		exitCh <- err
-	}()
-
-	return exitCh
-}
-
-func (k *Kafka) streamLogs(ctx context.Context) error {
-	if k.ErrorStream == nil && k.OutputStream == nil {
-		return nil
-	}
-
-	logOptions := docker.LogsOptions{
-		Context:      ctx,
-		Container:    k.ContainerID,
-		ErrorStream:  k.ErrorStream,
-		OutputStream: k.OutputStream,
-		Stderr:       k.ErrorStream != nil,
-		Stdout:       k.OutputStream != nil,
-		Follow:       true,
-	}
-	return k.Client.Logs(logOptions)
-}
-
-// Start starts the Kafka container using an ifrit runner
-func (k *Kafka) Start() error {
-	p := ifrit.Invoke(k)
-
-	select {
-	case <-p.Ready():
-		return nil
-	case err := <-p.Wait():
-		return err
-	}
-}
-
-// Stop stops and removes the Kafka container
-func (k *Kafka) Stop() error {
-	k.mutex.Lock()
-	if k.stopped {
-		k.mutex.Unlock()
-		return errors.Errorf("container %s already stopped", k.ContainerID)
-	}
-	k.stopped = true
-	k.mutex.Unlock()
-
-	return k.Client.StopContainer(k.ContainerID, 0)
-}
diff --git a/integration/nwo/runner/zookeeper.go b/integration/nwo/runner/zookeeper.go
deleted file mode 100644
index 0d5e946396b..00000000000
--- a/integration/nwo/runner/zookeeper.go
+++ /dev/null
@@ -1,247 +0,0 @@
-/*
-Copyright IBM Corp. All Rights Reserved.
-
-SPDX-License-Identifier: Apache-2.0
-*/
-
-package runner
-
-import (
-	"context"
-	"fmt"
-	"io"
-	"net"
-	"os"
-	"sync"
-	"time"
-
-	docker "github.com/fsouza/go-dockerclient"
-	"github.com/pkg/errors"
-	"github.com/tedsuo/ifrit"
-)
-
-const ZooKeeperDefaultImage = "confluentinc/cp-zookeeper:5.3.1"
-
-type ZooKeeper struct {
-	Client         *docker.Client
-	Image          string
-	HostIP         string
-	HostPort       []int
-	ContainerPorts []docker.Port
-	Name           string
-	StartTimeout   time.Duration
-
-	NetworkName string
-	ClientPort  docker.Port
-	LeaderPort  docker.Port
-	PeerPort    docker.Port
-	ZooMyID     int
-	ZooServers  string
-
-	ErrorStream  io.Writer
-	OutputStream io.Writer
-
-	containerID      string
-	containerAddress string
-	address          string
-
-	mutex   sync.Mutex
-	stopped bool
-}
-
-func (z *ZooKeeper) Run(sigCh <-chan os.Signal, ready chan<- struct{}) error {
-	if z.Image == "" {
-		z.Image = ZooKeeperDefaultImage
-	}
-
-	if z.Name == "" {
-		z.Name = DefaultNamer()
-	}
-
-	if z.HostIP == "" {
-		z.HostIP = "127.0.0.1"
-	}
-
-	if z.ContainerPorts == nil {
-		if z.ClientPort == docker.Port("") {
-			z.ClientPort = docker.Port("2181/tcp")
-		}
-		if z.LeaderPort == docker.Port("") {
-			z.LeaderPort = docker.Port("3888/tcp")
-		}
-		if z.PeerPort == docker.Port("") {
-			z.PeerPort = docker.Port("2888/tcp")
-		}
-
-		z.ContainerPorts = []docker.Port{
-			z.ClientPort,
-			z.LeaderPort,
-			z.PeerPort,
-		}
-	}
-
-	if z.StartTimeout == 0 {
-		z.StartTimeout = DefaultStartTimeout
-	}
-
-	if z.ZooMyID == 0 {
-		z.ZooMyID = 1
-	}
-
-	if z.Client == nil {
-		client, err := docker.NewClientFromEnv()
-		if err != nil {
-			return err
-		}
-		z.Client = client
-	}
-
-	containerOptions := docker.CreateContainerOptions{
-		Name: z.Name,
-		HostConfig: &docker.HostConfig{
-			AutoRemove: true,
-		},
-		Config: &docker.Config{
-			Image: z.Image,
-			Env: []string{
-				fmt.Sprintf("ZOOKEEPER_MY_ID=%d", z.ZooMyID),
-				fmt.Sprintf("ZOOKEEPER_SERVERS=%s", z.ZooServers),
-				fmt.Sprintf("ZOOKEEPER_CLIENT_PORT=%s", z.ClientPort.Port()),
-			},
-		},
-	}
-
-	if z.NetworkName != "" {
-		nw, err := z.Client.NetworkInfo(z.NetworkName)
-		if err != nil {
-			return err
-		}
-
-		containerOptions.NetworkingConfig = &docker.NetworkingConfig{
-			EndpointsConfig: map[string]*docker.EndpointConfig{
-				z.NetworkName: {
-					NetworkID: nw.ID,
-				},
-			},
-		}
-	}
-
-	container, err := z.Client.CreateContainer(containerOptions)
-	if err != nil {
-		return err
-	}
-	z.containerID = container.ID
-
-	err = z.Client.StartContainer(container.ID, nil)
-	if err != nil {
-		return err
-	}
-	defer z.Stop()
-
-	container, err = z.Client.InspectContainer(container.ID)
-	if err != nil {
-		return err
-	}
-
-	z.containerAddress = net.JoinHostPort(
-		container.NetworkSettings.IPAddress,
-		z.ContainerPorts[0].Port(),
-	)
-
-	streamCtx, streamCancel := context.WithCancel(context.Background())
-	defer streamCancel()
-	go z.streamLogs(streamCtx)
-
-	containerExit := z.wait()
-	ctx, cancel := context.WithTimeout(context.Background(), z.StartTimeout)
-	defer cancel()
-
-	select {
-	case <-ctx.Done():
-		return errors.Wrapf(ctx.Err(), "zookeeper in container %s did not start", z.containerID)
-	case <-containerExit:
-		return errors.New("container exited before ready")
-	default:
-		z.address = z.containerAddress
-	}
-
-	close(ready)
-
-	for {
-		select {
-		case err := <-containerExit:
-			return err
-		case <-sigCh:
-			if err := z.Stop(); err != nil {
-				return err
-			}
-		}
-	}
-}
-
-func (z *ZooKeeper) wait() <-chan error {
-	exitCh := make(chan error)
-	go func() {
-		exitCode, err := z.Client.WaitContainer(z.containerID)
-		if err == nil {
-			err = fmt.Errorf("zookeeper: process exited with %d", exitCode)
-		}
-		exitCh <- err
-	}()
-
-	return exitCh
-}
-
-func (z *ZooKeeper) streamLogs(ctx context.Context) error {
-	if z.ErrorStream == nil && z.OutputStream == nil {
-		return nil
-	}
-
-	logOptions := docker.LogsOptions{
-		Context:      ctx,
-		Container:    z.ContainerID(),
-		ErrorStream:  z.ErrorStream,
-		OutputStream: z.OutputStream,
-		Stderr:       z.ErrorStream != nil,
-		Stdout:       z.OutputStream != nil,
-		Follow:       true,
-	}
-	return z.Client.Logs(logOptions)
-}
-
-func (z *ZooKeeper) ContainerID() string {
-	return z.containerID
-}
-
-func (z *ZooKeeper) ContainerAddress() string {
-	return z.containerAddress
-}
-
-func (z *ZooKeeper) Start() error {
-	p := ifrit.Invoke(z)
-
-	select {
-	case <-p.Ready():
-		return nil
-	case err := <-p.Wait():
-		return err
-	}
-}
-
-func (z *ZooKeeper) Stop() error {
-	z.mutex.Lock()
-	if z.stopped {
-		z.mutex.Unlock()
-		return errors.Errorf("container %s already stopped", z.Name)
-	}
-	z.stopped = true
-	z.mutex.Unlock()
-
-	err := z.Client.StopContainer(z.containerID, 0)
-	if err != nil {
-		return err
-	}
-
-	_, err = z.Client.PruneVolumes(docker.PruneVolumesOptions{})
-	return err
-}
diff --git a/integration/nwo/standard_networks.go b/integration/nwo/standard_networks.go
index 70ab1da1b1a..a75e8d71325 100644
--- a/integration/nwo/standard_networks.go
+++ b/integration/nwo/standard_networks.go
@@ -181,16 +181,6 @@ func MultiChannelBasicSolo() *Config {
 	return config
 }
 
-func BasicKafka() *Config {
-	config := BasicSolo()
-
-	config.Consensus.Type = "kafka"
-	config.Consensus.ZooKeepers = 1
-	config.Consensus.Brokers = 1
-
-	return config
-}
-
 func BasicEtcdRaft() *Config {
 	config := BasicSolo()
 
diff --git a/integration/nwo/template/configtx_template.go b/integration/nwo/template/configtx_template.go
index 409b740e2a2..25192270742 100644
--- a/integration/nwo/template/configtx_template.go
+++ b/integration/nwo/template/configtx_template.go
@@ -151,12 +151,6 @@ Profiles:{{ range .Profiles }}
         PreferredMaxBytes: 512 KB
       Capabilities:
         V2_0: true
-      {{- if eq $w.Consensus.Type "kafka" }}
-      Kafka:
-        Brokers:{{ range $w.BrokerAddresses "HostPort" }}
-        - {{ . }}
-        {{- end }}
-      {{- end }}
       {{- if eq $w.Consensus.Type "etcdraft" }}
       EtcdRaft:
         Options:
diff --git a/integration/nwo/template/orderer_template.go b/integration/nwo/template/orderer_template.go
index 65b9f8248f0..efbbdc50a90 100644
--- a/integration/nwo/template/orderer_template.go
+++ b/integration/nwo/template/orderer_template.go
@@ -53,38 +53,6 @@ General:
     TimeWindow: 15m
 FileLedger:
   Location: {{ .OrdererDir Orderer }}/system
-{{ if eq .Consensus.Type "kafka" -}}
-Kafka:
-  Retry:
-    ShortInterval: 5s
-    ShortTotal: 10m
-    LongInterval: 5m
-    LongTotal: 12h
-    NetworkTimeouts:
-      DialTimeout: 10s
-      ReadTimeout: 10s
-      WriteTimeout: 10s
-    Metadata:
-      RetryBackoff: 250ms
-      RetryMax: 3
-    Producer:
-      RetryBackoff: 100ms
-      RetryMax: 3
-    Consumer:
-      RetryBackoff: 2s
-  Topic:
-    ReplicationFactor: 1
-  Verbose: false
-  TLS:
-    Enabled: false
-    PrivateKey:
-    Certificate:
-    RootCAs:
-  SASLPlain:
-    Enabled: false
-    User:
-    Password:
-  Version:{{ end }}
 Debug:
   BroadcastTraceDir:
   DeliverTraceDir:
diff --git a/integration/pvtdata/pvtdata_test.go b/integration/pvtdata/pvtdata_test.go
index 1012ebfb17f..584766f7e85 100644
--- a/integration/pvtdata/pvtdata_test.go
+++ b/integration/pvtdata/pvtdata_test.go
@@ -182,7 +182,6 @@ var _ bool = Describe("PrivateData", func() {
 			network.Bootstrap()
 
 			members := grouper.Members{
-				{Name: "brokers", Runner: network.BrokerGroupRunner()},
 				{Name: "orderers", Runner: network.OrdererGroupRunner()},
 			}
 			networkRunner := grouper.NewOrdered(syscall.SIGTERM, members)
diff --git a/integration/raft/migration_test.go b/integration/raft/migration_test.go
index f9b7148eec9..cd7f42b6685 100644
--- a/integration/raft/migration_test.go
+++ b/integration/raft/migration_test.go
@@ -40,7 +40,6 @@ var _ = Describe("Solo2RaftMigration", func() {
 		network *nwo.Network
 
 		process                ifrit.Process
-		brokerProc             ifrit.Process
 		o1Proc, o2Proc, o3Proc ifrit.Process
 
 		o1Runner, o2Runner *ginkgomon.Runner
@@ -68,11 +67,6 @@ var _ = Describe("Solo2RaftMigration", func() {
 			}
 		}
 
-		if brokerProc != nil {
-			brokerProc.Signal(syscall.SIGTERM)
-			Eventually(brokerProc.Wait(), network.EventuallyTimeout).Should(Receive())
-		}
-
 		if network != nil {
 			network.Cleanup()
 		}