Cannot start Indexing Task.(Cannot find instance of indexer to talk)

Hi!

I start Coordinator note(I am using Derby as metadata storage), after Historical Node then Overlord. All starting without error.

After I POST a task, the task log is:


2016-01-18T12:45:28,059 INFO [main] io.druid.indexing.common.actions.RemoteTaskActionClient - Performing action for task[index_hadoop_healthcare_2016-01-18T12:45:22.731Z]: LockTryAcquireAction{interval=2010-08-01T00:00:00.000Z/2015-09-01T00:00:00.000Z}
2016-01-18T12:45:28,067 ERROR [main] io.druid.curator.discovery.ServerDiscoverySelector - No server instance found
2016-01-18T12:45:28,068 WARN [main] io.druid.indexing.common.actions.RemoteTaskActionClient - Exception submitting action for task[index_hadoop_healthcare_2016-01-18T12:45:22.731Z]
java.io.IOException: Failed to locate service uri
at io.druid.indexing.common.actions.RemoteTaskActionClient.submit(RemoteTaskActionClient.java:91) [druid-indexing-service-0.8.2.jar:0.8.2]
at io.druid.indexing.common.task.HadoopIndexTask.isReady(HadoopIndexTask.java:135) [druid-indexing-service-0.8.2.jar:0.8.2]
at io.druid.indexing.worker.executor.ExecutorLifecycle.start(ExecutorLifecycle.java:120) [druid-indexing-service-0.8.2.jar:0.8.2]
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:1.8.0_66]
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:1.8.0_66]
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_66]
at java.lang.reflect.Method.invoke(Method.java:497) ~[?:1.8.0_66]
at com.metamx.common.lifecycle.Lifecycle$AnnotationBasedHandler.start(Lifecycle.java:325) [java-util-0.27.4.jar:?]
at com.metamx.common.lifecycle.Lifecycle.addMaybeStartHandler(Lifecycle.java:230) [java-util-0.27.4.jar:?]
at com.metamx.common.lifecycle.Lifecycle.addMaybeStartManagedInstance(Lifecycle.java:176) [java-util-0.27.4.jar:?]
at io.druid.guice.LifecycleScope$1.get(LifecycleScope.java:76) [druid-api-0.3.13.jar:0.8.2]
at com.google.inject.internal.InternalFactoryToProviderAdapter.get(InternalFactoryToProviderAdapter.java:41) [guice-4.0-beta.jar:?]
at com.google.inject.internal.InjectorImpl$3$1.call(InjectorImpl.java:1005) [guice-4.0-beta.jar:?]
at com.google.inject.internal.InjectorImpl.callInContext(InjectorImpl.java:1051) [guice-4.0-beta.jar:?]
at com.google.inject.internal.InjectorImpl$3.get(InjectorImpl.java:1001) [guice-4.0-beta.jar:?]
at com.google.inject.internal.InjectorImpl.getInstance(InjectorImpl.java:1040) [guice-4.0-beta.jar:?]
at io.druid.cli.CliPeon.run(CliPeon.java:225) [druid-services-0.8.2.jar:0.8.2]
at io.druid.cli.Main.main(Main.java:91) [druid-services-0.8.2.jar:0.8.2]
Caused by: com.metamx.common.ISE: Cannot find instance of indexer to talk to!
at io.druid.indexing.common.actions.RemoteTaskActionClient.getServiceInstance(RemoteTaskActionClient.java:165) ~[druid-indexing-service-0.8.2.jar:0.8.2]
at io.druid.indexing.common.actions.RemoteTaskActionClient.submit(RemoteTaskActionClient.java:86) ~[druid-indexing-service-0.8.2.jar:0.8.2]
… 17 more
2016-01-18T12:45:28,075 INFO [main] io.druid.indexing.common.actions.RemoteTaskActionClient - Will try again in [PT66.530S].
2016-01-18T12:46:34,605 ERROR [main] io.druid.curator.discovery.ServerDiscoverySelector - No server instance found

``


Here is the part of my _common/common.runtime.properties: (which is untouched)
# Indexing service discovery
druid.selectors.indexing.serviceName=overlord


Here is the part of my overlord/runtime.properties: (which is untouched)
#druid.host=localhost
#druid.port=8090
druid.service=overlord


After I POST the task first overlord logs about it, then Coordinator logs about it.

Thanks for future solutions.

Kind regards.

It seems that the property druid.selectors.indexing.serviceName=overlord is not being passed properly to the task peon which is causing this.
Can you make sure you have included common.runtime.properties while starting the overlord ?

you can also try adding druid.indexer.fork.property.druid.selectors.indexing.serviceName=overlord in the runtime.properties to make sure its passed properly.

Thank you Nishant, I add the config/_common to the classpath and now it is working.

Hi,
I have the same issue.

I checked every config was fine.

Hi Chanh, can you please include your runtime.properties of your various nodes?

,Hi

I’m having the same issue and I added druid.indexer.fork.property.druid.selectors.indexing.serviceName to the middle manager runtime.properties and the _common is in the class pat

Any help will be great

!Thanks

בתאריך יום שני, 18 בינואר 2016 בשעה 15:07:04 UTC+2, מאת prinull:

hi guys i hit this issue a few days ago and i did a root cause analysis and detailed a work around in my blog. hope it helps

https://mash213.wordpress.com/2018/10/02/druid-indexing-fails-ise-cannot-find-instance-of-indexer-to-talk-to/