You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@quickstep.apache.org by "Zuyu Zhang (JIRA)" <ji...@apache.org> on 2017/02/05 23:11:41 UTC

[jira] [Created] (QUICKSTEP-74) TextScanWorkOrder executed against HDFS

Zuyu Zhang created QUICKSTEP-74:
-----------------------------------

             Summary: TextScanWorkOrder executed against HDFS
                 Key: QUICKSTEP-74
                 URL: https://issues.apache.org/jira/browse/QUICKSTEP-74
             Project: Apache Quickstep
          Issue Type: Improvement
          Components: Relational Operators
            Reporter: Zuyu Zhang


Currently, TextScanWorkOrder uses C library APIs to process a file for bulk loading.

In the distributed version, we may store Quickstep blocks in HDFS. Note that data exchanger could not pull a remote block if it is not loaded in the memory of a peer node. Thus, we still need HDFS as the ground truth, and bulk loaded blocks need to store in HDFS.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)