HIVE-29330: HiveMetaStoreAuthorizer creating new HiveConf per thread …

[dengzhhu653]

…increases overhead

What changes were proposed in this pull request?

Why are the changes needed?

HiveMetaStoreAuthorizer can act as the pre event listener inside the HMS, or as the metadata filter hook on the client. If it's insides the HMS, getConf() should have loaded all properties in hive-site.xml, otherwise it contains the information to talk with the HMS at least, as the call is from client as a filter hook.

We don't need to create a new HiveConf per thread, it's wasteful and might spend dozens of milliseconds to load the properties, log the warn message "HiveConf of name xx does not exist" continuously if the property is only specified in MetastoreConf.

Does this PR introduce any user-facing change?

No

How was this patch tested?

The warn message of HiveConf doesn't exist is gone in an affected HMS

[sonarqubecloud]

Quality Gate passed

Issues
7 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[wecharyu]

There are other places where creating new HiveConf may not be necessary. For example:

hive/ql/src/java/org/apache/hadoop/hive/ql/ddl/table/execute/AlterTableExecuteAnalyzer.java

Lines 146 to 154 in d9ec041

	private static AlterTableExecuteDesc getExpireSnapshotDesc(TableName tableName, Map<String, String> partitionSpec,
	List<Node> children, HiveConf conf) throws SemanticException {
	AlterTableExecuteSpec<ExpireSnapshotsSpec> spec;
	if (children.size() == 1) {
	spec = new AlterTableExecuteSpec(EXPIRE_SNAPSHOT, null);
	return new AlterTableExecuteDesc(tableName, partitionSpec, spec);
	}
	ZoneId timeZone = SessionState.get() == null ?
	new HiveConf().getLocalTimeZone() :

[wecharyu]

nit: unnecessary empty line.

[wecharyu]

How about making it public and use it for a clone?