[
https://issues.apache.org/jira/browse/HIVE-27150?focusedWorklogId=853049&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-853049
]
ASF GitHub Bot logged work on HIVE-27150:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 27/Mar/23 06:08
Start Date: 27/Mar/23 06:08
Worklog Time Spent: 10m
Work Description: saihemanth-cloudera commented on code in PR #4123:
URL: https://github.com/apache/hive/pull/4123#discussion_r1148824344
##########
standalone-metastore/metastore-server/src/main/java/org/apache/hadoop/hive/metastore/ObjectStore.java:
##########
@@ -3101,6 +3100,22 @@ public boolean dropPartition(String catName, String
dbName, String tableName,
return success;
}
+ @Override
+ public boolean dropPartition(String catName, String dbName, String
tableName, String partName)
+ throws MetaException, NoSuchObjectException, InvalidObjectException,
InvalidInputException {
+ boolean success = false;
+ try {
+ openTransaction();
+ dropPartitionsInternal(catName, dbName, tableName,
Arrays.asList(partName), true, true);
Review Comment:
I don't think this would improve the performance by any means. Consider
dropping 10k partitions, each partition would have to access same number of
tables in the underlying db to update the records, so it makes sense to batch
them and implement with direct SQL. But for a single partition since it'll
access same number of tables in the DB, I don't think it'll make sense to
implement this feature.
For example, [HIVE-26035](https://issues.apache.org/jira/browse/HIVE-26035)
(see the details in the jira) proved that implementing direct SQL actually
improved the performance by running against benchmark tests.
Similarly can you provide any evidence that this patch also has an edge by
running those tests? (Probably you might have to add some tests(e.g: dropping
10K+ single partitions).
Issue Time Tracking
-------------------
Worklog Id: (was: 853049)
Time Spent: 1h 20m (was: 1h 10m)
> Drop single partition can also support direct sql
> -------------------------------------------------
>
> Key: HIVE-27150
> URL: https://issues.apache.org/jira/browse/HIVE-27150
> Project: Hive
> Issue Type: Improvement
> Components: Hive
> Reporter: Wechar
> Assignee: Wechar
> Priority: Major
> Labels: pull-request-available
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> *Background:*
> [HIVE-6980|https://issues.apache.org/jira/browse/HIVE-6980] supports direct
> sql for drop_partitions, we can reuse this huge improvement in drop_partition.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
