Rajesh Balamohan created HIVE-24368:
---------------------------------------
Summary: Optimise AcidUtils::getAcidFilesForStats for ACID tables
Key: HIVE-24368
URL: https://issues.apache.org/jira/browse/HIVE-24368
Project: Hive
Issue Type: Improvement
Components: HiveServer2
Reporter: Rajesh Balamohan
After insert, hive gathers statistics for ACID table and that becomes expensive
over time, due to number of delta folders and scanning .
[https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/io/AcidUtils.java#L2648]
{noformat}
public static List<FileStatus> getAcidFilesForStats(
Table table, Path dir, Configuration jc, FileSystem fs) throws
IOException {
...
Directory acidInfo = AcidUtils.getAcidState(fs, dir, jc, idList, null,
false, hdfsDirSnapshots);
...
..+ other calls
...
}
{noformat}
Runtime keeps increasing as more deltas are generated.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
