[DRILL-3907] Enable ScanBatch to provide a merge on pre-sorted readers - ASF JIRA

Attach files

Attach Screenshot

Add vote

Voters

Watch issue

Watchers

Create sub-task

Link

Clone

Update Comment Author

Replace String in Comment

Update Comment Visibility

Delete Comments

XML

Word

Printable

JSON

Details

Type: Bug
Status: Open
Priority: Major
Resolution: Unresolved
Affects Version/s: None
Fix Version/s: Future
Component/s: Storage - Other
Labels:
None

Description

In some situations, individual record readers will be presorted by a key. Given Drill's approach to parallelization, it is possible that a Single ScanBatch will be interacting with multiple readers. If we want to maintain the collation of the underlying data, we need to Drill to do a n-way merge on the streams as they are read into Drill. This functionality already exists in the MergingReceiver.

This JIRA is to refactor merging receiver so that the underlying N-Way merge of batches can be used in other locations. We then need to decide whether to incorporate it directly into the ScanBatch (when needed) or to do something external. We also need to resolve how we decide whether the collation that could be provided by utilizing an n-way merge is necessary (to avoid paying the cost of maintaining an unused collation).