Hello, We have a topic with 3 partitions where 1.5gb data getting pushed per minute but the whole data getting pushed to one partition(0) and other 2 partitions(1,2) not getting any data. I believe Kafka by default take cares of distributing data between brokers and partitions. Please help out with your thoughts. Thanks, Arshad ... View more

Hi All, We have druid on our hdp cluster hdfs as deep storage and set of nodes being used for that druid cluster. Is it possible to have another separate druid cluster with another set of nodes in same hdp cluster ? If you have different nodes for brokers/middle/history/coordination/overlord nodes for new druid cluster in same hdp cluster is it possible to have 2 druid clusters isolated in same hdp and ambari. My question comes with zookeeper leader election and how middle manager nodes will route the segments will point to particular hdfs deep storage. Please help. Thanks, Arshad ... View more

Hi, I have created a simple workflow which imports data from Oracle to hive using sqoop and it is running fine. As part of testing multi tenancy/new namespace, I have put the workflow.xml, coordinator.xml in new namespace and changed namenode to pointing to new namepsace tried running the workflow using oozie. Workflow fails with error JA017 could not look up launched Hadoop job I'd. Could any one please advice what could be wrong. Thanks in advance. Thanks, Arshad ... View more

We have added a new namespace to our existing one namespace cluster. We have services like hive/spark/hbase/Kafka/ which are using default namespace for storage and processing. We would like to have different hive metastore for our new namespace. Is it possible to have different hive metastore for different namespaces ? I see that we can give new namespace as a location when you are creating hive tables but does any one implemented federation/ multi tenant for hbase? I think federation is not applicable for Kafka as it stores the streams in the host where Kafka brokers are running not in hdfs. I really appreciate if some one can guide me with the above questions. Thanks, Arshad ... View more

We have added a new namespace to our existing one namespace cluster. We have services like hive/spark/hbase/Kafka/ which are using default namespace for storage and processing. We would like to have different hive metastore for our new namespace. Is it possible to have different hive metastore for different namespaces ? I see that we can give new namespace as a location when you are creating hive tables but does any one implemented federation/ multi tenant for hbase? I think federation is not applicable for Kafka as it stores the streams in the host where Kafka brokers are running not in hdfs. I really appreciate if some one can guide me with the above questions. Thanks, Arshad Thanks, Atehad ... View more

Hi, We have recently added a new namespace to existing hdp cluster using ambari. Since we have only one ambari we will have same configurations and properties for both the tenants or namespaces . Can we have different configurations and properties set at service level like hive for different namespaces ? Can we have different ambari UI for each namespace so we can have the different configurations and properties ? ... View more

I am trying to setup multi tenancy cluster for HDP and HDF. I dont find any documentation from hortonworks. Can any one share any documentation/links which i can refer ? ... View more