Unable to connect to some clusters (strimzi provisioned)

[chris-nolan]

Hi,

I've configured kafka-ui to point at four internal kafka clusters, two of these connect fine but the other two won't connect. It isn't a networking issue (kafka-ui is running locally, all clusters accessible from kadeck on the same desktop, DNS resolution fine). The two clusters it can't connect to are provisioned using Strimzi on k8s, but don't see how that can be an issue. On start up I get a few of the following WARN messages.

2022-11-28 07:50:19,699 WARN  [parallel-4] o.a.k.c.u.AppInfoParser: Error registering AppInfo mbean
javax.management.InstanceAlreadyExistsException: kafka.admin.client:type=app-info,id=kafka-ui-app
        at java.management/com.sun.jmx.mbeanserver.Repository.addMBean(Repository.java:436)
        at java.management/com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.registerWithRepository(DefaultMBeanServerInterceptor.java:1865)
        at java.management/com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.registerDynamicMBean(DefaultMBeanServerInterceptor.java:960)
        at java.management/com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.registerObject(DefaultMBeanServerInterceptor.java:895)
        at java.management/com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.registerMBean(DefaultMBeanServerInterceptor.java:320)
        at java.management/com.sun.jmx.mbeanserver.JmxMBeanServer.registerMBean(JmxMBeanServer.java:523)
        at org.apache.kafka.common.utils.AppInfoParser.registerAppInfo(AppInfoParser.java:64)
        at org.apache.kafka.clients.admin.KafkaAdminClient.<init>(KafkaAdminClient.java:604)
        at org.apache.kafka.clients.admin.KafkaAdminClient.createInternal(KafkaAdminClient.java:546)
        at org.apache.kafka.clients.admin.KafkaAdminClient.createInternal(KafkaAdminClient.java:485)
        at org.apache.kafka.clients.admin.Admin.create(Admin.java:134)
        at org.apache.kafka.clients.admin.AdminClient.create(AdminClient.java:39)
        at com.provectus.kafka.ui.service.AdminClientServiceImpl.lambda$createAdminClient$2(AdminClientServiceImpl.java:41)
        at reactor.core.publisher.MonoSupplier.call(MonoSupplier.java:86)
        at reactor.core.publisher.FluxFlatMap.trySubscribeScalarMap(FluxFlatMap.java:127)
        at reactor.core.publisher.MonoFlatMap.subscribeOrReturn(MonoFlatMap.java:53)
        at reactor.core.publisher.Mono.subscribe(Mono.java:4440)
        at reactor.core.publisher.FluxSwitchIfEmpty$SwitchIfEmptySubscriber.onComplete(FluxSwitchIfEmpty.java:82)
        at reactor.core.publisher.Operators.complete(Operators.java:137)
        at reactor.core.publisher.MonoEmpty.subscribe(MonoEmpty.java:46)
        at reactor.core.publisher.Mono.subscribe(Mono.java:4455)
        at reactor.core.publisher.FluxFlatMap$FlatMapMain.onNext(FluxFlatMap.java:426)
        at reactor.core.publisher.FluxPublishOn$PublishOnSubscriber.runAsync(FluxPublishOn.java:440)
        at reactor.core.publisher.FluxPublishOn$PublishOnSubscriber.run(FluxPublishOn.java:527)
        at reactor.core.scheduler.WorkerTask.call(WorkerTask.java:84)
        at reactor.core.scheduler.WorkerTask.call(WorkerTask.java:37)
        at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264)
        at java.base/java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:304)
        at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136)
        at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635)
        at java.base/java.lang.Thread.run(Thread.java:833)

I've triple checked the configuration which looks fine. Any ideas? Can provide more info if needed/run extra tests.

Cheers!

[Haarolean]

Hi, obviously we'd need your configuration to help you :)

[chris-nolan]

The docker image is built from provectuslabs/kafka-ui:latest and just copies in the fat jar for the customer deserializers

[Haarolean]

kafka-ui:latest points to latest release image (which was published in May) which doesn't have the custom serde feature yet. You should try master-labeled image instead.
A few other suggestions:

1. why don't you use docker-compose/application config with such a configuration? This command is almost unreadable
2. You don't have a volume binds for the files you reference via serde.x.filePath

javax.management.InstanceAlreadyExistsException: kafka.admin.client:type=app-info,id=kafka-ui-app
This probably means that you already have one instance of kafka-ui running. This is why I recommend using docker-compose once again, it tracks the state of your compose and won't create an extra instances for each run.

[chris-nolan]

Ok cool. I'm going to skip directly to deploying into kube with the helm chart. That is how we intend to deploy so I think its worth doing that now, just to make sure nothing weird is happening with my, yeah very unreadable setup.

[chris-nolan]

Deployed on kube using the helm chart and everything is working now!

[Haarolean]

@chris-nolan glad to hear!