How to read hadoop tables concurrently from multiple instances of a service?

13 views Asked by At

There is a hadoop table having thousands of records. I have a micro-service which has multiple instances and each instance will be connected to this hadoop table and should be able to read the data concurrently. I want to process the data from hadoop parallely so that no two service instances are processing the same data from the table. But, if there is a failure while processing the service should be able to convey the failure back to the table so that that particular data set can be given to some other service instance for processing (basically like a object lock system, 1 instance fetches few records from the table by applying a lock). What options of storage layer will support this requirement? Impala + Hadoop? or something else?

0

There are 0 answers