[
https://issues.apache.org/jira/browse/DRILL-6032?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16344403#comment-16344403
]
ASF GitHub Bot commented on DRILL-6032:
---------------------------------------
Github user paul-rogers commented on a diff in the pull request:
https://github.com/apache/drill/pull/1101#discussion_r164624503
--- Diff:
exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/aggregate/HashAggTemplate.java
---
@@ -215,6 +206,7 @@ public BatchHolder() {
MaterializedField outputField = materializedValueFields[i];
// Create a type-specific ValueVector for this value
vector = TypeHelper.getNewVector(outputField, allocator);
+ int columnSize = new RecordBatchSizer.ColumnSize(vector).estSize;
--- End diff --
Agree that the "sizer" should return the size of a fixed-width column even
if there are 0 rows. Thanks for catching and fixing this. Though, I will point
out that if you get a 0-length batch, just throw it away and read another; no
useful purpose is served by doing anything with an empty batch (except to pass
it along for schema purposes if this is the first batch: the so-called "fast
schema" that is supposed to work, but seems to not be fully implemented...)
I've found that Drill behaves poorly and erratically when trying to
allocate zero-sized vectors. Best not to do that. If your batch size is 0,
don't allocate vectors would be my general advice to avoid lots of grief.
> Use RecordBatchSizer to estimate size of columns in HashAgg
> -----------------------------------------------------------
>
> Key: DRILL-6032
> URL: https://issues.apache.org/jira/browse/DRILL-6032
> Project: Apache Drill
> Issue Type: Improvement
> Reporter: Timothy Farkas
> Assignee: Timothy Farkas
> Priority: Major
> Fix For: 1.13.0
>
>
> We need to use the RecordBatchSize to estimate the size of columns in the
> Partition batches created by HashAgg.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
