**overlord on server1 and its config is as follows **
The exception means that Tranquility can’t communicate with the indexing task to send data to it. It looks like you set the overlord’s druid.service=overlord, in which case you need to make sure your config for Tranquility has druid.selectors.indexing.serviceName=overlord. Also make sure your druid.discovery.curator.path is the same for Tranquility and the rest of your Druid nodes.
Your last statement seems to indicate that you don’t have any middle managers running. In general you should have at least one middle manager to use the indexing service (you can use druid.indexer.runner.type=local without middle managers for testing, but production systems should use druid.indexer.runner.type=remote with middle managers).
what i’ve done now is as follows,
overlord config is (on server No.1)
and one more thing is there
in my tranquility config i am setting druid.selectors.indexing.serviceName=overlord and when i submit/run it it shows like as follows :
Your value for druid.discovery.curator.path looks suspect. You set it to ‘/data/realtime/basePersist’ in your Tranquility spec, did you also set it to this value in your common.runtime.properties? If you didn’t, the default value is ‘/druid/discovery’ and you should set this value in your Tranquility spec.
As for your questions about the config for the realtime indexing task, the serviceName for the task will not (and should not) be the same as the serviceName you provide for your overlord, and Tranquility will parse your event using whatever timestampSpec you provide (i.e. format: yyyy-MM-dd HH:mm:ss) but it looks like it sends it to Druid as millis so you shouldn’t be concerned about that.
thanks for help… i changed druid.discovery.curator.path to /druid/discovery and it worked … thanks again