Important

You are viewing documentation for an older version of Confluent Platform. For the latest, click here.

Monitoring Kafka streaming ETL deployments

This demo shows users how to deploy a Kafka streaming ETL using KSQL for stream processing and Confluent Control Center for monitoring. All the components in the Confluent platform have security enabled end-to-end.

Overview

The use case is a streaming ETL deployment for real-time edits to real Wikipedia pages. Wikimedia Foundation has IRC channels that publish edits happening to real wiki pages (e.g. #en.wikipedia, #en.wiktionary) in real time. Using Kafka Connect, a Kafka source connector kafka-connect-irc streams raw messages from these IRC channels, and a custom Kafka Connect transform kafka-connect-transform-wikiedit transforms these messages and then the messages are written to a Kafka cluster. This demo uses KSQL for data enrichment, or you can optionally develop and run your own Kafka Streams application. Then a Kafka sink connector kafka-connect-elasticsearch streams the data out of Kafka, applying another custom Kafka Connect transform called NullFilter. The data is materialized into Elasticsearch for analysis by Kibana.

image

Note

This is a Docker environment and has all services running on one host. Do not use this demo in production. It is meant exclusively to easily demo the Confluent Platform. In production, Confluent Control Center should be deployed with a valid license and with its own dedicated metrics cluster, separate from the cluster with production traffic. Using a dedicated metrics cluster is more resilient because it continues to provide system health monitoring even if the production traffic cluster experiences issues.

Run demo

Follow along with the Demo 1: Install + Run video.

Demo validated with:

  • Docker version 17.06.1-ce
  • Docker Compose version 1.14.0 with Docker Compose file format 2.1
  • Java version 1.8.0_92
  • MacOS 10.12
  • git
  • jq

  1. Clone the cp-demo GitHub repository:

    $ git clone https://github.com/confluentinc/cp-demo

  2. In Docker’s advanced settings, increase the memory dedicated to Docker to at least 8GB (default is 2GB).

  3. From the cp-demo directory, start the entire demo by running a single command that generates the keys and certificates, brings up the Docker containers, and configures and validates the environment. This will take less than 5 minutes to complete.

    $ ./scripts/start.sh

  4. Use Google Chrome to view the Confluent Control Center GUI at http://localhost:9021. Click on the top right button that shows the current date, and change Last 4 hours to Last 30 minutes.

  5. View the data in the Kibana dashboard at http://localhost:5601/app/kibana#/dashboard/Wikipedia

Playbook

Confluent Control Center

Follow along with the Demo 2: Tour video.

  1. Monitoring –> System Health: Confluent Control Center landing page shows the overall system health of a given Kafka cluster. For capacity planning activities, view cluster utilization:

    • CPU: look at network and thread pool usage, produce and fetch request latencies
    • Network utilization: look at throughput per broker or per cluster
    • Disk utilization: look at disk space used by all log segments, per broker
    ../../../_images/landing_page.png

  2. Management –> Kafka Connect: Confluent Control Center uses the Kafka Connect API to manage Kafka connectors.

    • Kafka Connect Sources tab shows the connectors wikipedia-irc and replicate-topic. Click Edit to see the details of the connector configuration and custom transforms.

      image

    • Kafka Connect Sinks tab shows the connector elasticsearch-ksql. Click Edit to see the details of the connector configuration and custom transforms.

      image

  3. Monitoring –> Data Streams –> Message Delivery: hover over any chart to see number of messages and average latency within a minute time interval.

    image

    The Kafka Connect sink connectors have corresponding consumer groups connect-elasticsearch-ksql and connect-replicator. These consumer groups will be in the consumer group statistics in the stream monitoring charts.

    image

  4. Management –> Topics –> Topic Information: For a given topic, click on the three dots ... next to the topic name and click on View details. View which brokers are leaders for which partitions and the number of consumer groups currently consuming from this topic. Click on the boxed consumer group count to select a consumer group for which to monitor its data streams and jump to it.

    image

  5. Monitoring –> System Health: to identify bottlenecks, you can see a breakdown of produce and fetch latencies through the entire request lifecycle. Click on the line graph in the Request latency chart. The request latency values can be shown at the median, 95th, 99th, or 99.9th percentile. Depending on where the bottlenecks are, you can tune your brokers and clients appropriately.

    image

  6. Management -> Topics: click the + Create button on the top right to create a new topic in your Kafka cluster. You can also view and edit settings of Kafka topics in the cluster. Read more on Confluent Control Center topic management.

    image

KSQL

Follow along with the Demo 3: KSQL video.

In this demo, KSQL is authenticated and authorized to connect to the secured Kafka cluster, and it is already running queries as defined in the KSQL command file.

  1. The KSQL server is listening on port 8088. You have two options for interfacing with KSQL:

    1. Run KSQL CLI to get to the KSQL CLI prompt.

      $ docker-compose exec ksql-cli ksql http://localhost:8088

    2. Run the preview KSQL web interface. Navigate your browser to http://localhost:8088/index.html

  2. At the KSQL prompt, view the configured KSQL properties that were set with the KSQL properties file.

    ksql> SHOW PROPERTIES;

  3. View the existing KSQL streams and describe one of those streams called WIKIPEDIABOT.

    ksql> SHOW STREAMS;

 Stream Name              | Kafka Topic              | Format
--------------------------------------------------------------
 EN_WIKIPEDIA_GT_1_COUNTS | EN_WIKIPEDIA_GT_1_COUNTS | AVRO
 WIKIPEDIA                | wikipedia.parsed         | AVRO
 WIKIPEDIABOT             | WIKIPEDIABOT             | AVRO
 WIKIPEDIANOBOT           | WIKIPEDIANOBOT           | AVRO
 EN_WIKIPEDIA_GT_1_STREAM | EN_WIKIPEDIA_GT_1        | AVRO
--------------------------------------------------------------


ksql> DESCRIBE WIKIPEDIABOT;

 Field         | Type
-------------------------------------------
 ROWTIME       | BIGINT           (system)
 ROWKEY        | VARCHAR(STRING)  (system)
 CREATEDAT     | BIGINT
 WIKIPAGE      | VARCHAR(STRING)
 CHANNEL       | VARCHAR(STRING)
 USERNAME      | VARCHAR(STRING)
 COMMITMESSAGE | VARCHAR(STRING)
 BYTECHANGE    | INTEGER
 DIFFURL       | VARCHAR(STRING)
 ISNEW         | BOOLEAN
 ISMINOR       | BOOLEAN
 ISBOT         | BOOLEAN
 ISUNPATROLLED | BOOLEAN
-------------------------------------------

  4. View the existing KSQL tables and describe one of those tables called EN_WIKIPEDIA_GT_1.

    ksql> SHOW TABLES;

 Table Name        | Kafka Topic       | Format | Windowed
-----------------------------------------------------------
 EN_WIKIPEDIA_GT_1 | EN_WIKIPEDIA_GT_1 | AVRO   | true
-----------------------------------------------------------


ksql> DESCRIBE EN_WIKIPEDIA_GT_1;

 Field    | Type
--------------------------------------
 ROWTIME  | BIGINT           (system)
 ROWKEY   | VARCHAR(STRING)  (system)
 USERNAME | VARCHAR(STRING)  (key)
 WIKIPAGE | VARCHAR(STRING)  (key)
 COUNT    | BIGINT
--------------------------------------

  5. View the existing KSQL queries, which are continuously running, and explain one of those queries called CSAS_WIKIPEDIABOT.

    ksql> SHOW QUERIES;

 Query ID                      | Kafka Topic              | Query String
--------------------------------------------------------------------------------------------------
 CSAS_WIKIPEDIABOT             | WIKIPEDIABOT             | CREATE STREAM wikipediabot WITH (PARTITIONS=2,REPLICAS=2) AS SELECT * FROM wikipedia WHERE isbot = true;
 CTAS_EN_WIKIPEDIA_GT_1        | EN_WIKIPEDIA_GT_1        | CREATE TABLE en_wikipedia_gt_1 WITH (PARTITIONS=2,REPLICAS=2) AS SELECT username, wikipage, count(*) AS COUNT FROM wikipedia WINDOW TUMBLING (size 300 second) WHERE channel = '#en.wikipedia' GROUP BY username, wikipage HAVING count(*) > 1;
 CSAS_WIKIPEDIANOBOT           | WIKIPEDIANOBOT           | CREATE STREAM wikipedianobot WITH (PARTITIONS=2,REPLICAS=2) AS SELECT * FROM wikipedia WHERE isbot <> true;
 CSAS_EN_WIKIPEDIA_GT_1_COUNTS | EN_WIKIPEDIA_GT_1_COUNTS | CREATE STREAM en_wikipedia_gt_1_counts WITH (PARTITIONS=2,REPLICAS=2) AS SELECT * FROM en_wikipedia_gt_1_stream where ROWTIME is not null;
--------------------------------------------------------------------------------------------------


ksql> EXPLAIN CSAS_WIKIPEDIABOT;

Type                 : QUERY
SQL                  : CREATE STREAM wikipediabot WITH (PARTITIONS=2,REPLICAS=2) AS SELECT * FROM wikipedia WHERE isbot = true;


Local runtime statistics
------------------------
messages-per-sec:      1.07   total-messages:      1210     last-message: 2/16/18 4:47:16 PM UTC
 failed-messages:         0 failed-messages-per-sec:         0      last-failed:       n/a
(Statistics of the local KSQL server interaction with the Kafka topic WIKIPEDIABOT)

  6. At the KSQL prompt, view three messages from different KSQL streams and tables.

    ksql> SELECT * FROM WIKIPEDIABOT LIMIT 3;
ksql> SELECT * FROM EN_WIKIPEDIA_GT_1 LIMIT 3;
ksql> SELECT * FROM EN_WIKIPEDIA_GT_1_COUNTS LIMIT 3;

  7. In this demo, KSQL is run with Confluent Monitoring Interceptors configured which enables Confluent Control Center Data Streams to monitor KSQL queries. The consumer group names ksql_query_ correlate to the KSQL query names above, and Confluent Control Center is showing the records that are incoming to each query.

  • View throughput and latency of the incoming records for the persistent KSQL “Create Stream As Select” query CSAS_WIKIPEDIABOT, which is displayed as ksql_query_CSAS_WIKIPEDIABOT in Confluent Control Center.

    image

  • View throughput and latency of the incoming records for the persistent KSQL “Create Table As Select” query CTAS_EN_WIKIPEDIA_GT_1, which is displayed as ksql_query_CTAS_EN_WIKIPEDIA_GT_1 in Confluent Control Center.

    image

  • View throughput and latency of the incoming records for the persistent KSQL “Create Stream As Select” query CTAS_EN_WIKIPEDIA_GT_1_COUNTS, which is displayed as ksql_query_CSAS_EN_WIKIPEDIA_GT_1_COUNTS in Confluent Control Center.

    image

    Note

    In Confluent Control Center the stream monitoring graphs for consumer groups ksql_query_CSAS_EN_WIKIPEDIA_GT_1_COUNTS and EN_WIKIPEDIA_GT_1_COUNTS-consumer are displaying data at 5-minute intervals instead of smoothly like the other consumer groups. This is because Confluent Control Center displays data based on message timestamps, and the incoming stream for these consumer groups is a tumbling window with a window size of 5 minutes. Thus all its messages are timestamped to the beginning of each 5-minute window. This is also why the latency for these streams appears to be high. Kafka streaming tumbling windows are working as designed, and Confluent Control Center is reporting them accurately.

  1. This demo creates two streams EN_WIKIPEDIA_GT_1 and EN_WIKIPEDIA_GT_1_COUNTS, and the reason is to demonstrate how KSQL windows work. EN_WIKIPEDIA_GT_1 counts occurences with a tumbling window, and for a given key it writes a null into the table on the first seen message. The underlying Kafka topic for EN_WIKIPEDIA_GT_1 does not filter out those nulls, but since we want to send downstream just the counts greater than one, there is a separate Kafka topic for ``EN_WIKIPEDIA_GT_1_COUNTS which does filter out those nulls (e.g., the query has a clause where ROWTIME is not null). From the bash prompt, view those underlying Kafka topics.

    $ docker exec connect kafka-avro-console-consumer --bootstrap-server kafka1:9091 --topic EN_WIKIPEDIA_GT_1 \
  --property schema.registry.url=https://schemaregistry:8085 \
  --consumer.config /etc/kafka/secrets/client_without_interceptors.config --max-messages 10
null
{"USERNAME":"Atsme","WIKIPAGE":"Wikipedia:Articles for deletion/Metallurg Bratsk","COUNT":2}
null
null
null
{"USERNAME":"7.61.29.178","WIKIPAGE":"Tandem language learning","COUNT":2}
{"USERNAME":"Attar-Aram syria","WIKIPAGE":"Antiochus X Eusebes","COUNT":2}
...

$ docker exec connect kafka-avro-console-consumer --bootstrap-server kafka1:9091 --topic EN_WIKIPEDIA_GT_1_COUNTS \
  --property schema.registry.url=https://schemaregistry:8085 \
  --consumer.config /etc/kafka/secrets/client_without_interceptors.config --max-messages 10
{"USERNAME":"Atsme","COUNT":2,"WIKIPAGE":"Wikipedia:Articles for deletion/Metallurg Bratsk"}
{"USERNAME":"7.61.29.178","COUNT":2,"WIKIPAGE":"Tandem language learning"}
{"USERNAME":"Attar-Aram syria","COUNT":2,"WIKIPAGE":"Antiochus X Eusebes"}
{"USERNAME":"RonaldB","COUNT":2,"WIKIPAGE":"Wikipedia:Open proxy detection"}
{"USERNAME":"Dormskirk","COUNT":2,"WIKIPAGE":"Swindon Designer Outlet"}
{"USERNAME":"B.Bhargava Teja","COUNT":3,"WIKIPAGE":"Niluvu Dopidi"}
...

Consumer rebalances

Follow along with the Demo 4: Consumer Rebalances video.

Control Center shows which consumers in a consumer group are consuming from which partitions and on which brokers those partitions reside. Control Center updates as consumer rebalances occur in a consumer group.

  1. Start consuming from topic wikipedia.parsed with a new consumer group app with one consumer consumer_app_1. It will run in the background.

    $ ./scripts/app/start_consumer_app.sh 1

  2. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption. Click on the box View Details above the bar graph to drill down into consumer group details. This consumer group app has a single consumer consumer_app_1 consuming all of the partitions in the topic wikipedia.parsed. The first bar may be red because the consumer started in the middle of a time window and did not receive all messages produced during that window. This does not mean messages were lost.

    image

  3. Add a second consumer consumer_app_2 to the existing consumer group app.

    $ ./scripts/app/start_consumer_app.sh 2

  4. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption. Notice that the consumers consumer_app_1 and consumer_app_2 now share consumption of the partitions in the topic wikipedia.parsed. When the second consumer was added, that bar may be red for both consumers because a consumer rebalance occurred during that time window. This does not mean messages were lost, as you can confirm at the consumer group level.

    image

Slow consumers

Follow along with the Demo 5: Slow Consumers video.

Streams monitoring in Control Center can highlight consumers that are slow to keep up with the producers. This is critial to monitor for real-time applications where consumers should consume produced messages with as low latency as possible. To simulate a slow consumer, we will use Kafka’s quota feature to rate-limit consumption from the broker side, for just one of two consumers in a consumer group.

  1. Click on Data streams, and View Details for the consumer group app. Click on the left-hand blue circle on the consumption line to verify there are two consumers consumer_app_1 and consumer_app_2, that were created in an earlier section. If these two consumers are not running, start them as described in the section consumer rebalances.

  2. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption.

  3. Add a consumption quota for one of the consumers in the consumer group app.

    $ ./scripts/app/throttle_consumer.sh 1 add

    Note

    You are running a Docker demo environment with all services running on one host, which you would never do in production. Depending on your system resource availability, sometimes applying the quota may stall the consumer (KAFKA-5871), thus you may need to adjust the quota rate. See the ./scripts/app/throttle_consumer.sh script for syntax on modifying the quota rate.

    • If consumer group app does not increase latency, decrease the quota rate
    • If consumer group app seems to stall, increase the quota rate

  4. View the details of the consumer group app again, consumer_app_1 now shows high latency, and consumer_app_2 shows normal latency.

    image

  5. In the System Health dashboard, you see that the fetch request latency has likewise increased. This is the because the broker that has the partition that consumer_app_1 is consuming from is taking longer to service requests.

    image

  6. Click on the fetch request latency line graph to see a breakdown of produce and fetch latencies through the entire request lifecycle. The middle number does not necessarily equal the sum of the percentiles of individual segments because it is the total percentile latency.

    image

  7. Remove the consumption quota for the consumer. Latency for consumer_app_1 recovers to steady state values.

    $ ./scripts/app/throttle_consumer.sh 1 delete

Over consumption

Follow along with the Demo 6: Over Consumption video.

Streams monitoring in Control Center can highlight consumers that are over consuming some messages, which is an indication that consumers are processing a set of messages more than once. This may happen intentionally, for example an application with a software bug consumed and processed Kafka messages incorrectly, got a fix, and then reprocesses previous messages correctly. This may also happen unintentionally if an application crashes before committing processed messages. To simulate over consumption, we will use Kafka’s consumer offset reset tool to set the offset of the consumer group app to an earlier offset, thereby forcing the consumer group to reconsume messages it has previously read.

  1. Click on Data streams, and View Details for the consumer group app. Click on the blue circle on the consumption line on the left to verify there are two consumers consumer_app_1 and consumer_app_2, that were created in an earlier section. If these two consumers are not running and were never started, start them as described in the section consumer rebalances.

    image

  2. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption.

  3. Stop the consumer group app to stop consuming from topic wikipedia.parsed. Note that the command below stops the consumers gracefully with kill -15, so the consumers follow the shutdown sequence.

    $ ./scripts/app/stop_consumer_app_group_graceful.sh

  4. Wait for 2 minutes to let messages continue to be written to the topics for a while, without being consumed by the consumer group app. Notice the red bar which highlights that during the time window when the consumer group was stopped, there were some messages produced but not consumed. These messages are not missing, they are just not consumed because the consumer group stopped.

    image

  5. Reset the offset of the consumer group app by shifting 200 offsets backwards. The offset reset tool must be run when the consumer is completely stopped. Offset values in output shown below will vary.

    $ docker-compose exec kafka1 kafka-consumer-groups \
    --reset-offsets --group app --shift-by -200 --bootstrap-server kafka1:10091 \
    --all-topics --execute

    Your output should resemble:

    TOPIC            PARTITION NEW-OFFSET
wikipedia.parsed 1         4071
wikipedia.parsed 0         7944

  6. Restart consuming from topic wikipedia.parsed with the consumer group app with two consumers.

    $ ./scripts/app/start_consumer_app.sh 1
$ ./scripts/app/start_consumer_app.sh 2

  7. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption. Notice several things:

    • Even though the consumer group app was not running for some of this time, all messages are shown as delivered. This is because all bars are time windows relative to produce timestamp.
    • For some time intervals, the the bars are red and consumption line is above expected consumption because some messages were consumed twice due to rewinding offsets.
    • The latency peaks and then gradually decreases, because this is also relative to the produce timestamp.
    image

Under consumption

Follow along with the Demo 7: Under Consumption video.

Streams monitoring in Control Center can highlight consumers that are under consuming some messages. This may happen intentionally when consumers stop and restart and operators change the consumer offsets to the latest offset. This avoids delay processing messages that were produced while the consumers were stopped, especially when they care about real-time. This may also happen unintentionally if a consumer is offline for longer than the log retention period, or if a producer is configured for acks=0 and a broker suddenly fails before having a chance to replicate data to other brokers. To simulate under consumption, we will use Kafka’s consumer offset reset tool to set the offset of the consumer group app to the latest offset, thereby skipping messages that will never be read.

  1. Click on Data Streams, and View Details for the consumer group app. Click on the blue circle on the consumption line on the left to verify there are two consumers consumer_app_1 and consumer_app_2, that were created in an earlier section. If these two consumers are not running and were never started, start them as described in the section consumer rebalances.

    image

  2. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption.

  3. Stop the consumer group app to stop consuming from topic wikipedia.parsed. Note that the command below stops the consumers ungracefully with kill -9, so the consumers did not follow the shutdown sequence.

    $ ./scripts/app/stop_consumer_app_group_ungraceful.sh

  4. Wait for 2 minutes to let messages continue to be written to the topics for a while, without being consumed by the consumer group app. Notice the red bar which highlights that during the time window when the consumer group was stopped, there were some messages produced but not consumed. These messages are not missing, they are just not consumed because the consumer group stopped.

    image

  5. Wait for another few minutes and notice that the bar graph changes and there is a herringbone pattern to indicate that perhaps the consumer group stopped ungracefully.

    image

  6. Reset the offset of the consumer group app by setting it to latest offset. The offset reset tool must be run when the consumer is completely stopped. Offset values in output shown below will vary.

    $ docker-compose exec kafka1 kafka-consumer-groups \
  --reset-offsets --group app --to-latest --bootstrap-server kafka1:10091 \
  --all-topics --execute

    Your output should resemble:

    TOPIC            PARTITION NEW-OFFSET
wikipedia.parsed 1         8601
wikipedia.parsed 0         15135

  7. Restart consuming from topic wikipedia.parsed with the consumer group app with two consumers.

    $ ./scripts/app/start_consumer_app.sh 1
$ ./scripts/app/start_consumer_app.sh 2

  8. Let this consumer group run for 2 minutes until Control Center stream monitoring shows the consumer group app with steady consumption. Notice that during the time period that the consumer group app was not running, no produced messages are shown as delivered.

    image

Failed broker

Follow along with the Demo 8: Failed Broker video.

To simulate a failed broker, stop the Docker container running one of the two Kafka brokers.

  1. Stop the Docker container running Kafka broker 2.

    $ docker-compose stop kafka2

  2. After a few minutes, observe the System Health shows the broker count has gone down from 2 to 1, and there are many under replicated partitions.

    image

  3. View topic details to see that there are out of sync replicas on broker 2.

    image

  4. Restart the Docker container running Kafka broker 2.

    $ docker-compose start kafka2

  5. After about a minute, observe the System Health view in Confluent Control Center. The broker count has recovered to 2, and the topic partitions are back to reporting no under replicated partitions.

    image

  6. Click on the broker count 2 inside the circle to view when the broker counts changed.

    image

Alerting

Follow along with the Demo 9: Alerting video.

There are many types of Control Center alerts and many ways to configure them. Use the Alerts management page to define triggers and actions, or click on a streams monitoring graph for consumer groups or topics to setup alerts from there.

  1. This demo already has pre-configured triggers and actions. View the Alerts Triggers screen, and click Edit against each trigger to see configuration details.

    • The trigger Under Replicated Partitions happens when a broker reports non-zero under replicated partitions, and it causes an action Email Administrator.
    • The trigger Consumption Difference happens when consumption difference for the Elasticsearch connector consumer group is greater than 0, and it causes an action Email Administrator.
    image

  2. If you followed the steps in the failed broker section, view the Alert history to see that the trigger Under Replicated Partitions happened and caused an alert when you stopped broker 2.

  3. You can also trigger the Consumption Difference trigger. In the Kafka Connect -> Sinks screen, edit the running Elasticsearch sink connector.

  4. In the Kafka Connect view, pause the Elasticsearch sink connector by pressing the pause icon in the top right. This will stop consumption for the related consumer group.

    image

  5. View the Alert history to see that this trigger happened and caused an alert.

    image

Replicator

Confluent Replicator copies data from a source Kafka cluster to a destination Kafka cluster. The source and destination clusters are typically different clusters, but in this demo, Replicator is doing intra-cluster replication, i.e., the source and destination Kafka clusters are the same. As with the rest of the components in the solution, Confluent Replicator is also configured with security.

  1. Monitoring –> Data Streams –> Message Delivery: monitor throughput and latency of Confluent Replicator in the Data streams monitoring view. Replicator is a Kafka Connect source connector and has a corresponding consumer group connect-replicator.

    image

  2. Management –> Topics: scroll down to view the topics called wikipedia.parsed (Replicator is consuming data from this topic) and wikipedia.parsed.replica (Replicator automatically created this topic and is copying data to it). Click on Consumer Groups for the topic wikipedia.parsed and observe that one of the consumer groups is called connect-replicator.

    ../../../_images/replicator_topic_info.png

  3. Management –> Kafka Connect: pause the Replicator connector by pressing the pause icon in the top right. This will stop consumption for the related consumer group.

    image

  4. Observe that the connect-replicator consumer group has stopped consumption.

    ../../../_images/replicator_streams_stopped.png

  5. Restart the Replicator connector.

  6. Observe that the connect-replicator consumer group has resumed consumption. Notice several things:

    • Even though the consumer group connect-replicator was not running for some of this time, all messages are shown as delivered. This is because all bars are time windows relative to produce timestamp.
    • The latency peaks and then gradually decreases, because this is also relative to the produce timestamp.

Security

Follow along with the Security video.

All the components in this demo are enabled with many security features:

  • SSL for encryption, except for ZooKeeper which does not support SSL
  • SASL/PLAIN for authentication, except for ZooKeeper
  • Authorization. If a resource has no associated ACLs, then users are not allowed to access the resource, except super users
  • HTTPS for Schema Registry

Note

This demo showcases a secure Confluent Platform for educational purposes and is not meant to be complete best practices. There are certain differences between what is shown in the demo and what you should do in production:

  • Each component should have its own username, instead of authenticating all users as client
  • Authorize users only for operations that they need, instead of making all of them super users
  • If the PLAINTEXT security protocol is used, these ANONYMOUS usernames should not be configured as super users
  • Consider not even opening the PLAINTEXT port if SSL or SASL_SSL are configured

Encryption & Authentication

Each broker has four listener ports:

  • PLAINTEXT port called PLAINTEXT for users with no security enabled
  • SSL port port called SSL for users with just SSL without SASL
  • SASL_SSL port called SASL_SSL for communication between services inside Docker containers
  • SASL_SSL port called SASL_SSL_HOST for communication between any potential services outside of Docker that communicate to the Docker containers
port kafka1 kafka2
PLAINTEXT 10091 10092
SSL 11091 11092
SASL_SSL 9091 9092
SASL_SSL_HOST 29091 29092

Authorization

All the brokers in this demo authenticate as broker, and all other components authenticate as client. Per the broker configuration parameter super.users, as it is set in this demo, the only users that can communicate with the cluster are those that authenticate as broker or client, or users that connect via the PLAINTEXT port (their username is ANONYMOUS). All other users are not authorized to communicate with the cluster.

  1. Verify the ports on which the Kafka brokers are listening with the following command, and they should match the table shown below:

    $ docker-compose logs kafka1 | grep "Registered broker 1"
$ docker-compose logs kafka2 | grep "Registered broker 2"

  2. This demo automatically generates simple SSL certificates and creates keystores, truststores, and secures them with a password. To communicate with the brokers, Kafka clients may use any of the ports on which the brokers are listening. To use a security-enabled port, they must specify security parameters for keystores, truststores, password, or authentication so the Kafka command line client tools pass the security configuration file with interceptors or without interceptors with these security parameters. As an example, to communicate with the Kafka cluster to view all the active consumer groups:

    1. Communicate with brokers via the PLAINTEXT port

      # PLAINTEXT port
$ docker-compose exec kafka1 kafka-consumer-groups --list --bootstrap-server kafka1:10091

    2. Communicate with brokers via the SASL_SSL port, and SASL_SSL parameters configured via the --command-config argument for command line tools or --consumer.config for kafka-console-consumer.

      # SASL_SSL port with SASL_SSL parameters
$ docker-compose exec kafka1 kafka-consumer-groups --list --bootstrap-server kafka1:9091 \
   --command-config /etc/kafka/secrets/client_without_interceptors.config

    3. If you try to communicate with brokers via the SASL_SSL port but don’t specify the SASL_SSL parameters, it will fail

      # SASL_SSL port without SASL_SSL parameters
$ docker-compose exec kafka1 kafka-consumer-groups --list --bootstrap-server kafka1:9091

      Your output should resemble:

      Error: Executing consumer group command failed due to Request
METADATA failed on brokers List(kafka1:9091 (id: -1 rack: null))

  3. Verify the super users are configured for the authenticated users broker, client, and unauthenticated PLAINTEXT.

    $ docker-compose logs kafka1 | grep SUPER_USERS

    Your output should resemble:

    KAFKA_SUPER_USERS=User:client;User:schemaregistry;User:broker;User:ANONYMOUS

  4. Verify that a user client which authenticates via SASL can consume messages from topic wikipedia.parsed:

    $ ./scripts/consumers/listen_wikipedia.parsed.sh SASL

  5. Verify that a user which authenticates via SSL cannot consume messages from topic wikipedia.parsed. It should fail with an exception.

    $ ./scripts/consumers/listen_wikipedia.parsed.sh SSL

    Your output should resemble:

    [2018-01-12 21:13:18,481] ERROR Unknown error when running consumer: (kafka.tools.ConsoleConsumer$)
org.apache.kafka.common.errors.TopicAuthorizationException: Not authorized to access topics: [wikipedia.parsed]

  6. Verify that the broker’s Authorizer logger logs the denial event. As shown in the log message, the user which authenticates via SSL has a username CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US, not just client.

    # Authorizer logger logs the denied operation
$ docker-compose logs kafka1 | grep kafka.authorizer.logger

    Your output should resemble:

    [2018-01-12 21:13:18,454] INFO Principal = User:CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US is Denied Operation = Describe from host = 172.23.0.7 on resource = Topic:wikipedia.parsed (kafka.authorizer.logger) [2018-01-12
21:13:18,464] INFO Principal = User:CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US is Denied Operation = Describe from host = 172.23.0.7 on resource = Group:test (kafka.authorizer.logger)

  7. Add an ACL that authorizes user CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US, and then view the updated ACL configuration.

    $ docker-compose exec connect /usr/bin/kafka-acls \
    --authorizer-properties zookeeper.connect=zookeeper:2181 \
    --add --topic wikipedia.parsed \
    --allow-principal User:CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US \
    --operation Read --group test

$ docker-compose exec connect /usr/bin/kafka-acls \
    --authorizer-properties zookeeper.connect=zookeeper:2181 \
    --list --topic wikipedia.parsed --group test

    Your output should resemble:

    Current ACLs for resource ``Topic:wikipedia.parsed``:
User:CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US has Allow permission for operations: Read from hosts: \*

Current ACLs for resource ``Group:test``:
User:CN=client,OU=TEST,O=CONFLUENT,L=PaloAlto,ST=Ca,C=US has Allow permission for operations: Read from hosts: \*

  8. Verify that the user which authenticates via SSL is now authorized and can successfully consume some messages from topic wikipedia.parsed.

    $ ./scripts/consumers/listen_wikipedia.parsed.sh SSL

Schema Registry and REST Proxy

The connectors used in this demo are configured to automatically read and write Avro-formatted data, leveraging the Confluent Schema Registry . The Confluent REST Proxy is running for optional client access.

  1. View the Schema Registry subjects for topics that have registered schemas for their keys and/or values. Notice the security arguments passed into the curl command which are required to interact with the Schema Registry, which is listening for HTTPS on port 8085.

    $ docker-compose exec restproxy curl -X GET --cert /etc/kafka/secrets/schemaregistry.certificate.pem --key /etc/kafka/secrets/schemaregistry.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt https://schemaregistry:8085/subjects | jq .

[
  "ksql_query_CTAS_EN_WIKIPEDIA_GT_1-KSQL_Agg_Query_1526914100640-changelog-value",
  "ksql_query_CTAS_EN_WIKIPEDIA_GT_1-KSQL_Agg_Query_1526914100640-repartition-value",
  "EN_WIKIPEDIA_GT_1_COUNTS-value",
  "WIKIPEDIABOT-value",
  "EN_WIKIPEDIA_GT_1-value",
  "WIKIPEDIANOBOT-value",
  "wikipedia.parsed-value"
]

  2. Register a new Avro schema (a record with two fields username and userid) into Confluent Schema Registry for the value of a new topic users. Note the schema id that it returns, in this case id is 6.

    $ docker-compose exec restproxy curl -X POST -H "Content-Type: application/vnd.schemaregistry.v1+json" --cert /etc/kafka/secrets/schemaregistry.certificate.pem --key /etc/kafka/secrets/schemaregistry.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt --data '{ "schema": "[ { \"type\":\"record\", \"name\":\"user\", \"fields\": [ {\"name\":\"userid\",\"type\":\"long\"}, {\"name\":\"username\",\"type\":\"string\"} ]} ]" }' https://schemaregistry:8085/subjects/users-value/versions | jq .

{
  "id": 6
}

  3. View the new schema for the subject users-value.

    $ docker-compose exec restproxy curl -X GET --cert /etc/kafka/secrets/schemaregistry.certificate.pem --key /etc/kafka/secrets/schemaregistry.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt https://schemaregistry:8085/subjects/users-value/versions/1 | jq .

{
  "subject": "users-value",
  "version": 1,
  "id": 6,
  "schema": "{\"type\":\"record\",\"name\":\"user\",\"fields\":[{\"name\":\"username\",\"type\":\"string\"},{\"name\":\"userid\",\"type\":\"long\"}]}"
}

  4. Use the REST Proxy, which is listening for HTTPS on port 8086, to produce a message to the topic users, referencing schema id 6.

    $ docker-compose exec restproxy curl -X POST -H "Content-Type: application/vnd.kafka.avro.v2+json" -H "Accept: application/vnd.kafka.v2+json" --cert /etc/kafka/secrets/restproxy.certificate.pem --key /etc/kafka/secrets/restproxy.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt --data '{"value_schema_id": 6, "records": [{"value": {"user":{"userid": 1, "username": "Bunny Smith"}}}]}' https://restproxy:8086/topics/users

{"offsets":[{"partition":1,"offset":0,"error_code":null,"error":null}],"key_schema_id":null,"value_schema_id":6}

  5. Use the REST Proxy to consume the above message from the topic users. This is a series of steps.

    # Create consumer instance my_avro_consumer
$ docker-compose exec restproxy curl -X POST -H "Content-Type: application/vnd.kafka.v2+json" --cert /etc/kafka/secrets/restproxy.certificate.pem --key /etc/kafka/secrets/restproxy.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt --data '{"name": "my_consumer_instance", "format": "avro", "auto.offset.reset": "earliest"}' https://restproxy:8086/consumers/my_avro_consumer

# Subscribe my_avro_consumer to the `users` topic
$ docker-compose exec restproxy curl -X POST -H "Content-Type: application/vnd.kafka.v2+json" --cert /etc/kafka/secrets/restproxy.certificate.pem --key /etc/kafka/secrets/restproxy.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt --data '{"topics":["users"]}' https://restproxy:8086/consumers/my_avro_consumer/instances/my_consumer_instance/subscription

# Get messages for my_avro_consumer subscriptions (you may need to repeat this command)
$ docker-compose exec restproxy curl -X GET -H "Accept: application/vnd.kafka.avro.v2+json" --cert /etc/kafka/secrets/restproxy.certificate.pem --key /etc/kafka/secrets/restproxy.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt https://restproxy:8086/consumers/my_avro_consumer/instances/my_consumer_instance/records

# Delete the consumer instance my_avro_consumer
$ docker-compose exec restproxy curl -X DELETE -H "Content-Type: application/vnd.kafka.v2+json" --cert /etc/kafka/secrets/restproxy.certificate.pem --key /etc/kafka/secrets/restproxy.key --tlsv1.2 --cacert /etc/kafka/secrets/snakeoil-ca-1.crt https://restproxy:8086/consumers/my_avro_consumer/instances/my_consumer_instance

Troubleshooting the demo

  1. Verify the status of the Docker containers show Up state, except for the kafka-client container which is expected to have Exit 0 state. If any containers are not up, verify in the advanced Docker preferences settings that the memory available to Docker is at least 8 GB (default is 2 GB).

    $ docker-compose ps

    Your output should resemble:

             Name                        Command               State                              Ports
------------------------------------------------------------------------------------------------------------------------------
connect                   /etc/confluent/docker/run        Up       0.0.0.0:8083->8083/tcp, 9092/tcp
control-center            /etc/confluent/docker/run        Up       0.0.0.0:9021->9021/tcp
elasticsearch             /bin/bash bin/es-docker          Up       0.0.0.0:9200->9200/tcp, 0.0.0.0:9300->9300/tcp
kafka-client              bash -c -a echo Waiting fo ...   Exit 0
kafka1                    /etc/confluent/docker/run        Up       0.0.0.0:29091->29091/tcp, 0.0.0.0:9091->9091/tcp, 9092/tcp
kafka2                    /etc/confluent/docker/run        Up       0.0.0.0:29092->29092/tcp, 0.0.0.0:9092->9092/tcp
kibana                    /bin/sh -c /usr/local/bin/ ...   Up       0.0.0.0:5601->5601/tcp
ksql-cli                  perl -e while(1){ sleep 99 ...   Up       0.0.0.0:9098->9098/tcp
replicator                tail -f /dev/null                Up      8083/tcp, 9092/tcp
restproxy                 /etc/confluent/docker/run        Up       0.0.0.0:8082->8082/tcp, 0.0.0.0:8086->8086/tcp
schemaregistry            /etc/confluent/docker/run        Up       8081/tcp, 0.0.0.0:8085->8085/tcp
zookeeper                 /etc/confluent/docker/run        Up       0.0.0.0:2181->2181/tcp, 2888/tcp, 3888/tcp

  2. To view sample messages for each topic, including wikipedia.parsed:

    $ ./scripts/consumers/listen.sh

  3. If the data streams monitoring appears to stop for the Kafka source connector, restart the connect container.

    $ docker-compose restart connect

Teardown

  1. Stop the consumer group app to stop consuming from topic wikipedia.parsed. Note that the command below stops the consumers gracefully with kill -15, so the consumers follow the shutdown sequence.

    $ ./scripts/app/stop_consumer_app_group_graceful.sh

  2. Stop the Docker demo, destroy all components and clear all Docker volumes.

    $ ./scripts/stop.sh