ksql-datagen exits. Can't connect to broker

[jtravan3]

Description I followed the steps listed here (https://docs.confluent.io/platform/current/quickstart/ce-docker-quickstart.html#) to start up a local Kafka instance on Docker. ksql-datagen exits after a couple of minutes due to no connection to the broker

Troubleshooting I think this may be related to issues #53, #51, #25, #24, #12, and #10. I'm seeing the same behavior.

1.) I run docker-compose up -d and I can see all of the containers up and running. 2.) http://locahost:9021 shows no response 2.) Then I run docker-compose ps and see that ksql-datagen has exited. 3.) Running docker logs ksql-datagen shows that it can't connect to the broker.

Connection to node -1 (broker/192.168.224.3:29092) could not be established. Broker may not be available.

I have made no changes to the docker-compose.yml

Environment

• GitHub branch: 6.2.0-post
• Operating System: MacOS Big Sur 11.5
• Version of Docker: 3.5.2
• Version of Docker Compose: 1.29.2

[OneCricketeer]

I'm not sure 192.168.x.x addresses would be correct within the Docker bridge, but did you allocate the 6GB of memory to Docker as that page suggests?

[jtravan3]

@OneCricketeer I ended up backing down to version 5.5.5 (I'm not sure if that has anything to do with it) and then allocating 8 GB of memory to Docker instead of 6. After that, I can run it successfully. I have not tried to do the same with 6.2.0. Occasionaly, ksql-datagen will exit unexpectedly and I have to restart it but I have been able to stumble forward.

[OneCricketeer]

I don't think I've personally tried starting all components recently, but it usually is a memory issue from what I've seen. If you don't need the REST Proxy or Control Center, for example while you use ksql, you may comment them out in the yaml

[agates4]

Clean install of docker. Clone of https://github.com/confluentinc/cp-all-in-one/blob/6.2.0-post/cp-all-in-one/docker-compose.yml, docker-compose up, and containers exit.

Memory allocated: 8gb

This goes beyond memory allocation - there is an issue with this docker compose config.

[OneCricketeer]

I just started that file (after removing REST Proxy and Connect services and references to them), and I get a page saying there's a healthy cluster on http://localhost:9021 , no containers have exited after 5 minutes

Mac: 11.5.2 Docker: 20.10.8 darwin/amd64

update Adding back Connect and REST Proxy still works; no exited containers

[agates4]

Thank you for the comment - taking a double look and will report back.

[agates4]

The broker fails, triggering failure of the other images.

Mac: 11.4 (and I updated to 11.5.2, it doesn't work) Docker: 20.10.8 M1 ARM docker version

Steps:

1. docker system prune
2. Clone https://raw.githubusercontent.com/confluentinc/cp-all-in-one/6.2.0-post/cp-all-in-one/docker-compose.yml
3. docker-compose up
4. Wait
5. See the following error in the broker container:

07 23:47:42,271] TRACE [Controller id=1 epoch=1] Received response UpdateMetadataResponseData(errorCode=0) for request UPDATE_METADATA with correlation id 10 sent to broker broker:29092 (id: 1 rack: null tags: []) (state.change.logger)

[2021-09-07 23:47:45,721] INFO Started o.e.j.s.ServletContextHandler@796a728c{/v1/metadata,null,AVAILABLE} (org.eclipse.jetty.server.handler.ContextHandler)

[2021-09-07 23:47:46,526] INFO Waiting for 8 seconds for metric reporter topic _confluent-telemetry-metrics to become available. (io.confluent.cruisecontrol.metricsreporter.ConfluentMetricsSamplerBase)

[2021-09-07 23:47:47,240] INFO Stopped o.e.j.s.ServletContextHandler@796a728c{/v1/metadata,null,STOPPED} (org.eclipse.jetty.server.handler.ContextHandler)

[2021-09-07 23:47:47,241] INFO node0 Stopped scavenging (org.eclipse.jetty.server.session)

[2021-09-07 23:47:47,286] WARN KafkaHttpServer transitioned from STARTING to FAILED.: org.glassfish.jersey.internal.spi.AutoDiscoverable: : java.net.MalformedURLException: no !/ found in url spec:file:/usr/share/java/confluent-security/schema-validator/jersey-common-2.34.jar!/META-INF/services/org.glassfish.jersey.internal.spi.AutoDiscoverable. (io.confluent.http.server.KafkaHttpServerImpl)

org.glassfish.jersey.internal.ServiceConfigurationError: org.glassfish.jersey.internal.spi.AutoDiscoverable: : java.net.MalformedURLException: no !/ found in url spec:file:/usr/share/java/confluent-security/schema-validator/jersey-common-2.34.jar!/META-INF/services/org.glassfish.jersey.internal.spi.AutoDiscoverable

[agates4]

I am getting the same issue with 6.1.0. I am now going to try 6.0.3

[agates4]

My docker resources

[OneCricketeer]

I don't know what HTTP servlet Kafka (rather "Confluent Server") is trying to run, but I don't think that's related to the actual container crashing

You should try the community one instead, which runs a less customized Kafka broker https://github.com/confluentinc/cp-all-in-one/blob/6.2.0-post/cp-all-in-one-community/docker-compose.yml

[agates4]

ah thank you - this seems to have gotten the Kafka broker working, but the schema registry is failing.

[2021-09-08 05:00:24,885] INFO [Schema registry clientId=sr-1, groupId=schema-registry] (Re-)joining group (org.apache.kafka.clients.consumer.internals.AbstractCoordinator)
[2021-09-08 05:00:24,905] INFO [Schema registry clientId=sr-1, groupId=schema-registry] Successfully joined group with generation Generation{generationId=3, memberId='sr-1-273e5715-2660-4128-a38b-79034cf4eecc', protocol='v0'} (org.apache.kafka.clients.consumer.internals.AbstractCoordinator)
[2021-09-08 05:00:24,980] INFO [Schema registry clientId=sr-1, groupId=schema-registry] Successfully synced group in generation Generation{generationId=3, memberId='sr-1-273e5715-2660-4128-a38b-79034cf4eecc', protocol='v0'} (org.apache.kafka.clients.consumer.internals.AbstractCoordinator)
[2021-09-08 05:00:24,994] INFO Finished rebalance with leader election result: Assignment{version=1, error=0, leader='sr-1-273e5715-2660-4128-a38b-79034cf4eecc', leaderIdentity=version=1,host=schema-registry,port=8081,scheme=http,leaderEligibility=true} (io.confluent.kafka.schemaregistry.leaderelector.kafka.KafkaGroupLeaderElector)
[2021-09-08 05:00:25,059] INFO Wait to catch up until the offset at 3 (io.confluent.kafka.schemaregistry.storage.KafkaStore)
[2021-09-08 05:00:25,068] INFO Reached offset at 3 (io.confluent.kafka.schemaregistry.storage.KafkaStore)
Exception in thread "main" java.lang.ExceptionInInitializerError
	at org.eclipse.jetty.http.MimeTypes.<clinit>(MimeTypes.java:175)
	at org.eclipse.jetty.server.handler.gzip.GzipHandler.<init>(GzipHandler.java:190)
	at io.confluent.rest.ApplicationServer.wrapWithGzipHandler(ApplicationServer.java:468)
	at io.confluent.rest.ApplicationServer.wrapWithGzipHandler(ApplicationServer.java:477)
	at io.confluent.rest.ApplicationServer.finalizeHandlerCollection(ApplicationServer.java:213)
	at io.confluent.rest.ApplicationServer.doStart(ApplicationServer.java:230)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:73)
	at io.confluent.kafka.schemaregistry.rest.SchemaRegistryMain.main(SchemaRegistryMain.java:43)
Caused by: java.nio.charset.IllegalCharsetNameException: l;charset=iso-8859-1
	at java.base/java.nio.charset.Charset.checkName(Charset.java:308)
	at java.base/java.nio.charset.Charset.lookup2(Charset.java:482)
	at java.base/java.nio.charset.Charset.lookup(Charset.java:462)
	at java.base/java.nio.charset.Charset.forName(Charset.java:526)
	at org.eclipse.jetty.http.MimeTypes$Type.<init>(MimeTypes.java:107)
	at org.eclipse.jetty.http.MimeTypes$Type.<clinit>(MimeTypes.java:67)
	... 8 more

this seems related to char errors I was seeing with the broker in my previous tests. maybe there is something going on here with the 6.x.x releases.

[agates4]

https://github.com/localstack/localstack/issues/4456#issuecomment-905904064

I see an issue with alpine linux JDK on M1 with an illegal charset - potential problem here could be afflicting confluent images

[OneCricketeer]

Think you said you were running darwin/amd64, not M1 ARM docker version... In any case, you can specify a platform to emulate amd64 Docker runtime, I believe.

Last I checked, Confluent Platform is not supported on ARM based systems

[agates4]

sorry about that - yeah I am running on m1. I will try specifying a platform

[agates4]

ok - eugenetea/schema-registry-arm64:latest is an image for schema registry that appears to work well. from this issue thread https://github.com/confluentinc/cp-docker-images/issues/718

I see an issue thread open about support for ARM here https://github.com/confluentinc/common-docker/issues/117 looks like all of my issues were ARM related. I'd be curious to see who else in this thread is on an ARM system.

[agates4]

to those looking for a working M1 ARM broker+zookeeper+schema-registry, here is my docker-compose.yml

---
version: '3.4'

services:
  broker:
    container_name: broker
    depends_on:
      - zookeeper
    environment:
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://broker:29092,PLAINTEXT_HOST://localhost:9092
      KAFKA_BROKER_ID: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      KAFKA_JMX_HOSTNAME: localhost
      KAFKA_JMX_PORT: 9101
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_TRANSACTION_STATE_LOG_MIN_ISR: 1
      KAFKA_TRANSACTION_STATE_LOG_REPLICATION_FACTOR: 1
      KAFKA_ZOOKEEPER_CONNECT: zookeeper:2181
    hostname: broker
    image: confluentinc/cp-kafka:6.2.0
    ports:
      - '29092:29092'
      - '9092:9092'
      - '9101:9101'
  schema-registry:
    container_name: schema-registry
    depends_on:
      - broker
    environment:
      SCHEMA_REGISTRY_HOST_NAME: schema-registry
      SCHEMA_REGISTRY_KAFKASTORE_BOOTSTRAP_SERVERS: broker:29092
      SCHEMA_REGISTRY_LISTENERS: http://0.0.0.0:8081
    hostname: schema-registry
    image: eugenetea/schema-registry-arm64:latest
    platform: linux/arm64/v8
    ports:
      - '8081:8081'
  zookeeper:
    container_name: zookeeper
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000
    hostname: zookeeper
    image: confluentinc/cp-zookeeper:6.2.0
    ports:
      - '2181:2181'

[nicolepastrana]

This is what worked for me in my M1 Chip:

• zookeeper image: confluentinc/cp-zookeeper:latest.arm64
• broker image: confluentinc/cp-server:latest.arm64
• schema-registry image: confluentinc/cp-schema-registry:latest.arm64
• connect image: cnfldemos/cp-server-connect-datagen:0.5.0-6.2.0 [I haven't tested with the latest cnfldemos/cp-server-connect-datagen:0.5.3-7.1.0]
• control-center image: confluentinc/cp-enterprise-control-center:latest.arm64
• ksqldb-server image: confluentinc/cp-ksqldb-server:latest.arm64
• ksqldb-cli image: confluentinc/cp-ksqldb-cli:latest.arm64
• ksql-datagen image: confluentinc/ksqldb-examples:latest.arm64
• rest-proxy image: confluentinc/cp-kafka-rest:latest.arm64

https://github.com/confluentinc/cp-all-in-one/blob/7.1.1-post/cp-all-in-one/docker-compose.yml