I am a newer to Druid. When I setup Druid cluster, I encountered a problem : Service[druid:overlord] call failed with status: 400 Bad Request.
The cluster structure:
node1 - run coordinaor process; overlord process; broker process;
node2 - run historical process; middleManager process; tranquility
node3 is same with node2.
Druid version: 0.9.1.1
tranquility version: 0.8.0
I want to read data from Kafka in realtime using tranquility. When started coordinator, overlord, broker, historical, middlemanager process, I want to start tranquility, suing command
“bin/tranquility kafka -configFile /home/druid-0.9.1.1/conf/tranquility/kafka-test.json”, then the error mesage occured, and the tranquility process exit.
The kafka-test.json file is in the attachment, and also the detail error message.
Anyone who can help me!
Yufeng Wang
error_message (89.3 KB)
kafka-test.json (2.33 KB)
Hi Yufeng,
Your logs show that Druid is having trouble creating a temporary directory:
2016-08-24 08:29:08,837 [finagle/netty3-1] ERROR com.metamx.common.scala.control$ - Non-retryable response for uri[/druid/indexer/v1/task] with content: {“error”:“Instantiation of [simple type, class io.druid.segment.indexing.RealtimeTuningConfig] value failed: Failed to create directory within 10000 attempts (tried 1472027349460-0 to 1472027349460-9999)”}
Make sure that java.io.tmpdir exists and that it’s writable by Druid. Take a look at ‘JVM Configuration Best Practices’ here: http://druid.io/docs/latest/configuration/index.html
Thank you David. I finally use Imply and it works. Imply is much more simple 
在 2016年8月25日星期四 UTC+8上午2:03:23，David Lim写道：
Hello,
I have the Druid on a server and I am trying to import through Imply, however it does not work, it does the import on the Druid included in Imply so create a new connection or modify the existing one.
How did you manage to load an external Druid from Imply?
Thanks for the help