You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@kylin.apache.org by li...@apache.org on 2024/04/01 05:35:58 UTC
(kylin) 07/14: KYLIN-5767 Calculate total rows wrongly when connecting jdbc datasource
This is an automated email from the ASF dual-hosted git repository.
liyang pushed a commit to branch kylin5
in repository https://gitbox.apache.org/repos/asf/kylin.git
commit d3b4469484a79102071db26114f6602b30d70f43
Author: Liang.Hua <36...@users.noreply.github.com>
AuthorDate: Fri Jul 21 16:31:39 2023 +0800
KYLIN-5767 Calculate total rows wrongly when connecting jdbc datasource
---------
Co-authored-by: liang.hua <li...@kyligence.io>
---
.../kylin/engine/spark/NSparkCubingEngine.java | 4 ++++
.../engine/spark/builder/SnapshotBuilder.scala | 7 +++++--
.../engine/spark/job/NSparkCubingJobTest.java | 8 ++++++++
.../kylin/source/jdbc/DefaultSourceConnector.java | 5 +++++
.../apache/kylin/source/jdbc/ISourceConnector.java | 3 +++
.../apache/kylin/source/jdbc/JdbcSourceInput.java | 11 ++++++++++
.../kylin/source/jdbc/JdbcSourceInputTest.java | 24 ++++++++++++++++++++++
7 files changed, 60 insertions(+), 2 deletions(-)
diff --git a/src/spark-project/engine-spark/src/main/java/org/apache/kylin/engine/spark/NSparkCubingEngine.java b/src/spark-project/engine-spark/src/main/java/org/apache/kylin/engine/spark/NSparkCubingEngine.java
index 00078a86ad..76a2bcb9ff 100644
--- a/src/spark-project/engine-spark/src/main/java/org/apache/kylin/engine/spark/NSparkCubingEngine.java
+++ b/src/spark-project/engine-spark/src/main/java/org/apache/kylin/engine/spark/NSparkCubingEngine.java
@@ -50,6 +50,10 @@ public class NSparkCubingEngine implements NCubingEngine {
* @return the Dataset<Row>, its schema consists of table column's name, for example, [column1,column2,column3]
*/
Dataset<Row> getSourceData(TableDesc table, SparkSession ss, Map<String, String> parameters);
+
+ default Long getSourceDataCount(TableDesc table, SparkSession ss, Map<String, String> parameters) {
+ return getSourceData(table, ss, parameters).count();
+ }
}
public interface NSparkCubingStorage {
diff --git a/src/spark-project/engine-spark/src/main/scala/org/apache/kylin/engine/spark/builder/SnapshotBuilder.scala b/src/spark-project/engine-spark/src/main/scala/org/apache/kylin/engine/spark/builder/SnapshotBuilder.scala
index e217a31c99..1a9ef63544 100644
--- a/src/spark-project/engine-spark/src/main/scala/org/apache/kylin/engine/spark/builder/SnapshotBuilder.scala
+++ b/src/spark-project/engine-spark/src/main/scala/org/apache/kylin/engine/spark/builder/SnapshotBuilder.scala
@@ -199,8 +199,11 @@ class SnapshotBuilder(var jobId: String) extends Logging with Serializable {
case e: Throwable => logWarning(s"Calculate table ${tableDesc.getIdentity}'s total rows exception", e)
}
logInfo(s"Calculate table ${tableDesc.getIdentity}'s total rows from source data")
- val sourceData = getSourceData(ss, tableDesc)
- val totalRows = sourceData.count()
+ val params = NProjectManager.getInstance(KylinConfig.getInstanceFromEnv)
+ .getProject(tableDesc.getProject).getLegalOverrideKylinProps
+ val totalRows = SourceFactory
+ .createEngineAdapter(tableDesc, classOf[NSparkCubingEngine.NSparkCubingSource])
+ .getSourceDataCount(tableDesc, ss, params)
logInfo(s"Table ${tableDesc.getIdentity}'s total rows is ${totalRows}'")
totalRows
}
diff --git a/src/spark-project/engine-spark/src/test/java/org/apache/kylin/engine/spark/job/NSparkCubingJobTest.java b/src/spark-project/engine-spark/src/test/java/org/apache/kylin/engine/spark/job/NSparkCubingJobTest.java
index b86eb333f3..2078e44056 100644
--- a/src/spark-project/engine-spark/src/test/java/org/apache/kylin/engine/spark/job/NSparkCubingJobTest.java
+++ b/src/spark-project/engine-spark/src/test/java/org/apache/kylin/engine/spark/job/NSparkCubingJobTest.java
@@ -163,6 +163,14 @@ public class NSparkCubingJobTest extends NLocalWithSparkSessionTest {
getLookTables(df).forEach(table -> Assert.assertNotNull(table.getLastSnapshotPath()));
}
+ @Test
+ public void testCalculateTableTotalRows() {
+ NTableMetadataManager tableMetadataManager = NTableMetadataManager.getInstance(config, getProject());
+ TableDesc tableDesc = tableMetadataManager.getTableDesc("DEFAULT.TEST_ORDER");
+ long totalRows = new SnapshotBuilder().calculateTableTotalRows(null, tableDesc, ss);
+ Assert.assertEquals(5000, totalRows);
+ }
+
private Set<TableDesc> getLookTables(NDataflow df) {
return df.getModel().getLookupTables().stream().map(TableRef::getTableDesc).collect(Collectors.toSet());
}
diff --git a/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/DefaultSourceConnector.java b/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/DefaultSourceConnector.java
index 1ea35704ed..1723810955 100644
--- a/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/DefaultSourceConnector.java
+++ b/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/DefaultSourceConnector.java
@@ -49,4 +49,9 @@ public class DefaultSourceConnector extends DefaultAdaptor implements ISourceCon
.option("driver", driver).option("query", sql).options(params).load();
}
+ @Override
+ public Dataset<Row> getCountData(KylinConfig kylinConfig, SparkSession sparkSession, String sql, Map<String, String> params) {
+ return getSourceData(kylinConfig, sparkSession, sql, params);
+ }
+
}
diff --git a/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/ISourceConnector.java b/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/ISourceConnector.java
index e2d91bc0e5..b4ab3b8b5a 100644
--- a/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/ISourceConnector.java
+++ b/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/ISourceConnector.java
@@ -27,4 +27,7 @@ import org.apache.spark.sql.SparkSession;
public interface ISourceConnector {
Dataset<Row> getSourceData(KylinConfig kylinConfig, SparkSession sparkSession, String sql,
Map<String, String> params);
+
+ Dataset<Row> getCountData(KylinConfig kylinConfig, SparkSession sparkSession, String sql,
+ Map<String, String> params);
}
diff --git a/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/JdbcSourceInput.java b/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/JdbcSourceInput.java
index b6d2cf8557..608e85fa24 100644
--- a/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/JdbcSourceInput.java
+++ b/src/spark-project/source-jdbc/src/main/java/org/apache/kylin/source/jdbc/JdbcSourceInput.java
@@ -62,4 +62,15 @@ public class JdbcSourceInput implements NSparkCubingEngine.NSparkCubingSource {
StructType sparkSchema = dataset.schema();
return dataset.select(SparderTypeUtil.alignDataTypeAndName(sparkSchema, kylinSchema));
}
+
+ @Override
+ public Long getSourceDataCount(TableDesc table, SparkSession ss, Map<String, String> parameters) {
+ String sql = String.format(Locale.ROOT, "select count(*) from %s", table.getIdentity());
+ KylinConfig config = table.getConfig();
+ ISourceConnector connector = (ISourceConnector) ClassUtil.newInstance(config.getJdbcSourceConnector());
+ parameters.put("table", table.getIdentity());
+ log.info("Use source connector: {}, sql: {} ", connector.getClass().getCanonicalName(), sql);
+ Dataset<Row> dataset = connector.getCountData(config, ss, sql, parameters);
+ return dataset.first().getLong(0);
+ }
}
diff --git a/src/spark-project/source-jdbc/src/test/java/org/apache/kylin/source/jdbc/JdbcSourceInputTest.java b/src/spark-project/source-jdbc/src/test/java/org/apache/kylin/source/jdbc/JdbcSourceInputTest.java
index 24e11d57a1..2b205c531b 100644
--- a/src/spark-project/source-jdbc/src/test/java/org/apache/kylin/source/jdbc/JdbcSourceInputTest.java
+++ b/src/spark-project/source-jdbc/src/test/java/org/apache/kylin/source/jdbc/JdbcSourceInputTest.java
@@ -29,6 +29,7 @@ import org.apache.kylin.metadata.model.NTableMetadataManager;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;
+import org.junit.Assert;
import org.junit.BeforeClass;
import org.junit.Test;
@@ -63,4 +64,27 @@ public class JdbcSourceInputTest extends JdbcTestBase {
Dataset<Row> sourceData = cubingSource.getSourceData(tableDesc, ss, Maps.newHashMap());
System.out.println(sourceData.schema());
}
+
+ @Test
+ public void testGetSourceDataCount() {
+ NTableMetadataManager tableMgr = NTableMetadataManager.getInstance(getTestConfig(), "ssb");
+ TableDesc tableDesc = tableMgr.getTableDesc("SSB.CUSTOMER");
+ ISource source = SourceFactory.getSource(new ISourceAware() {
+ @Override
+ public int getSourceType() {
+ return ISourceAware.ID_JDBC;
+ }
+
+ @Override
+ public KylinConfig getConfig() {
+ return getTestConfig();
+ }
+ });
+ NSparkCubingEngine.NSparkCubingSource cubingSource = source
+ .adaptToBuildEngine(NSparkCubingEngine.NSparkCubingSource.class);
+ Assert.assertTrue(cubingSource instanceof JdbcSourceInput);
+ Long countData = cubingSource.getSourceDataCount(tableDesc, ss, Maps.newHashMap());
+ Long expectedCount = cubingSource.getSourceData(tableDesc, ss, Maps.newHashMap()).count();
+ Assert.assertEquals(expectedCount, countData);
+ }
}