[
https://issues.apache.org/jira/browse/HIVE-25975?focusedWorklogId=734758&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-734758
]
ASF GitHub Bot logged work on HIVE-25975:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 01/Mar/22 16:41
Start Date: 01/Mar/22 16:41
Worklog Time Spent: 10m
Work Description: marton-bod commented on a change in pull request #3060:
URL: https://github.com/apache/hive/pull/3060#discussion_r816948553
##########
File path: ql/src/java/org/apache/hadoop/hive/ql/plan/DynamicPartitionCtx.java
##########
@@ -50,6 +52,14 @@
private String defaultPartName; // default partition name in case of null or
empty value
private int maxPartsPerNode; // maximum dynamic partitions created per
mapper/reducer
private Pattern whiteListPattern;
+ /**
+ * Expressions describing a custom way of sorting the table before write.
Expressions can reference simple
+ * column descriptions or a tree of expressions containing more columns and
UDFs.
+ * Can be useful for custom bucket/hash sorting.
+ * A custom expression should be a lambda that is given the original column
description expressions as per read
+ * schema and returns a single expression. Example for simply just
referencing column 3: cols -> cols.get(3).clone()
+ */
+ private transient List<Function<List<ExprNodeDesc>, ExprNodeDesc>>
customSortExpressions;
Review comment:
Yeah, I was wondering if it's okay if these expressions don't get to the
executors. But I got my answer - Thanks!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 734758)
Time Spent: 4h (was: 3h 50m)
> Optimize ClusteredWriter for bucketed Iceberg tables
> ----------------------------------------------------
>
> Key: HIVE-25975
> URL: https://issues.apache.org/jira/browse/HIVE-25975
> Project: Hive
> Issue Type: Improvement
> Reporter: Ádám Szita
> Assignee: Ádám Szita
> Priority: Major
> Labels: pull-request-available
> Time Spent: 4h
> Remaining Estimate: 0h
>
> The first version of the ClusteredWriter in Hive-Iceberg will be lenient for
> bucketed tables: i.e. the records do not need to be ordered by the bucket
> values, the writer will just close its current file and open a new one for
> out-of-order records.
> This is suboptimal for the long-term due to creating many small files. Spark
> uses a UDF to compute the bucket value for each record and therefore it is
> able to order the records by bucket values, achieving optimal clustering.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
