Error Cluster batch load run

HI Team ,

I a trying to load the batch file in druid 0.15.2 version ,

Why i need to place my batch file at both the server master and data server , than only I am able to ingest .

Please guide me . I need to place the file at one server only in this cluster set up ,

I would be great ful , if someone can reply or talk to me on this ,

Urgent help required .


Amit SHarma

Hi Amit,
Do you mean your input data file which you need to ingest into druid?

If yes, it can anywhere like HDFS or some NFS path or one of your local path on the node you are running your curl command to ingest via batch.

Are you able to do this sample tests?

Can we have a quick call ? If you have time

If you are running a clustered setup, the data must be accessible from both overlord and middle manager servers. You can try two options

  1. Put file in a common NFS mount between two systems or better yet, use S3.

  2. Run middle manager and overlord on same node. Though this may not be ideal for production setups, must be good enough for Dev.

Also, since you are using latest version, the new UI’s data loader might be a viable option too, though it’s a manual way of loading data