Skip to content

Commit

Permalink
Add Iceberg tests with MinIO
Browse files Browse the repository at this point in the history
  • Loading branch information
Arkadiusz Czajkowski committed Mar 2, 2022
1 parent a13dbbd commit 19e595f
Show file tree
Hide file tree
Showing 11 changed files with 533 additions and 103 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -64,7 +64,7 @@ public HiveMinioDataLake(String bucketName, Map<String, String> hiveHadoopFilesT
this.hiveHadoop = closer.register(
HiveHadoop.builder()
.withFilesToMount(ImmutableMap.<String, String>builder()
.put("hive_s3_insert_overwrite/hive-core-site.xml", "/etc/hadoop/conf/core-site.xml")
.put("hive_minio_datalake/hive-core-site.xml", "/etc/hadoop/conf/core-site.xml")
.putAll(hiveHadoopFilesToMount)
.buildOrThrow())
.withImage(hiveHadoopImage)
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -20,4 +20,8 @@
<name>fs.s3a.path.style.access</name>
<value>true</value>
</property>
<property>
<name>fs.s3.impl</name>
<value>org.apache.hadoop.fs.s3a.S3AFileSystem</value>
</property>
</configuration>
12 changes: 12 additions & 0 deletions plugin/trino-iceberg/pom.xml
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,13 @@
<properties>
<air.main.basedir>${project.parent.basedir}</air.main.basedir>
<dep.iceberg.version>0.13.1</dep.iceberg.version>
<!--
Project's default for air.test.parallel is 'methods'. By design, 'classes' makes TestNG run tests from one class in a single thread.
As side effect it prevents TestNG from initializing multiple test instances upfront, which happens with 'methods'.
A potential downside can be long tail single-threaded execution of a single long test class.
This change is due to MVN Surefire issue which is addressed here https://github.com/apache/maven-surefire/pull/407
-->
<air.test.parallel>classes</air.test.parallel>
</properties>

<dependencies>
Expand Down Expand Up @@ -84,6 +91,11 @@
<artifactId>units</artifactId>
</dependency>

<dependency>
<groupId>com.amazonaws</groupId>
<artifactId>aws-java-sdk-s3</artifactId>
</dependency>

<dependency>
<groupId>com.fasterxml.jackson.core</groupId>
<artifactId>jackson-core</artifactId>
Expand Down

Large diffs are not rendered by default.

Original file line number Diff line number Diff line change
@@ -0,0 +1,203 @@
/*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package io.trino.plugin.iceberg;

import com.google.common.collect.ImmutableList;
import com.google.common.collect.ImmutableMap;
import io.trino.Session;
import io.trino.plugin.hive.containers.HiveMinioDataLake;
import io.trino.testing.QueryRunner;
import org.testng.annotations.DataProvider;
import org.testng.annotations.Test;

import java.io.BufferedInputStream;
import java.util.ArrayList;
import java.util.List;

import static com.google.common.collect.ImmutableList.toImmutableList;
import static io.trino.plugin.hive.containers.HiveMinioDataLake.ACCESS_KEY;
import static io.trino.plugin.hive.containers.HiveMinioDataLake.SECRET_KEY;
import static io.trino.testing.sql.TestTable.randomTableSuffix;
import static java.lang.String.format;
import static org.assertj.core.api.Assertions.assertThat;

public abstract class BaseIcebergMinioConnectorTest
extends BaseIcebergConnectorTest
{
private static final List<String> TEST_REPARTITION_EXCLUSIONS = ImmutableList.of(
TEST_REPARTITION_COMPLEX,
TEST_REPARTITION_SAME_COLUMN_MULTIPLE_TIMES);

private final String bucketName;
private HiveMinioDataLake hiveMinioDataLake;

public BaseIcebergMinioConnectorTest(IcebergFileFormat format)
{
super(format);
this.bucketName = "test-iceberg-minio-integration-test-" + randomTableSuffix();
}

@Override
protected QueryRunner createQueryRunner()
throws Exception
{
this.hiveMinioDataLake = closeAfterClass(new HiveMinioDataLake(bucketName, ImmutableMap.of()));
this.hiveMinioDataLake.start();
return createQueryRunner(
ImmutableMap.<String, String>builder()
.put("iceberg.catalog.type", "HIVE_METASTORE")
.put("hive.metastore.uri", "thrift://" + hiveMinioDataLake.getHiveHadoop().getHiveMetastoreEndpoint())
.put("hive.s3.aws-access-key", ACCESS_KEY)
.put("hive.s3.aws-secret-key", SECRET_KEY)
.put("hive.s3.endpoint", "http://" + hiveMinioDataLake.getMinio().getMinioApiEndpoint())
.put("hive.s3.path-style-access", "true")
.put("hive.s3.streaming.part-size", "5MB")
.buildOrThrow());
}

@Override
protected SchemaInitializer.Builder createSchemaInitializer(String schemaName)
{
return super.createSchemaInitializer(schemaName)
.withSchemaProperties(ImmutableMap.of("location", "'s3://" + bucketName + "/" + schemaName + "'"));
}

@Override
protected String createSchemaSql(String schemaName)
{
return "CREATE SCHEMA IF NOT EXISTS " + schemaName + " WITH (location = 's3://" + bucketName + "/" + schemaName + "')";
}

@Override
protected String createNewTableLocationBase(String schemaName)
{
return "s3://" + bucketName + "/" + schemaName;
}

@Override
protected Long calculateFileSystemFileSize(String filePath)
{
return this.hiveMinioDataLake.getS3Client()
.getObject(bucketName, filePath.replace("s3://" + bucketName + "/", ""))
.getObjectMetadata()
.getContentLength();
}

@Override
protected BufferedInputStream createDataFileStreamForFileSystem(String filePath)
{
return new BufferedInputStream(this.hiveMinioDataLake.getS3Client()
.getObject(bucketName, filePath.replace("s3://" + bucketName + "/", "")).getObjectContent());
}

@Override
protected List<String> getAllDataFilesFromTableDirectory(String tableName)
{
String tableLocation = getSession().getSchema().orElseThrow() + "/" + tableName + "/data";
return this.hiveMinioDataLake.getS3Client().listObjects(bucketName, tableLocation)
.getObjectSummaries()
.stream()
.map(object -> "s3://" + bucketName + "/" + object.getKey())
.filter(object -> !object.endsWith(".crc"))
.collect(toImmutableList());
}

/**
* DataProvider overload to exclude 2 test cases from generic tests:
* - {@link BaseIcebergConnectorTest#testRepartitionDataOnCtas(Session, String, int)}
* - {@link BaseIcebergConnectorTest#testRepartitionDataOnInsert(Session, String, int)}
* <p>
* 1. partitioning -> 'bucket(custkey, 4)', 'truncate(comment, 1)'
* Test environment provides higher memory usage reaching over 1,5GB per test case.
* This leads to test flakiness when both test classes run in parallel and may consume
* over 3GB which is current hard limit for CI and may end up with OOM.
* <p>
* Limiting dataset to 300 rows in dedicated test methods for MinIO solves this issue:
* - {@link BaseIcebergMinioConnectorTest#testRepartitionOnMinio()} ()}
* <p>
* 2. partitioning -> 'truncate(comment, 1)', 'orderstatus', 'bucket(comment, 2)'
* Test environment causes HMS to consume more time during createTable operation.
* This leads to read timed out and retry operation ended with TableAlreadyExists
* (HMS managed to end operation on it's side). As such behaviour wasn't observed
* in normal circumstances, test was adopted to avoid it.
* <p>
* Limiting dataset to 300 rows in dedicated test methods for MinIO solves this issue:
* - {@link BaseIcebergMinioConnectorTest#testRepartitionOnMinio()} ()}
*/
@Override
@DataProvider
public Object[][] repartitioningDataProvider()
{
Object[][] defaultTestsData = super.repartitioningDataProvider();
List<Object[]> minioTestData = new ArrayList<>();
for (Object[] testData : defaultTestsData) {
if (!TEST_REPARTITION_EXCLUSIONS.contains(testData[1])) {
minioTestData.add(testData);
}
}
return minioTestData.toArray(new Object[minioTestData.size()][]);
}

@Test
public void testRepartitionOnMinio()
{
String sourceQuery = "SELECT * FROM tpch.tiny.orders ORDER BY orderkey LIMIT 300";
// complex; would exceed 100 open writers limit in IcebergPageSink without write repartitioning
testRepartitionData(getSession(), sourceQuery, true, TEST_REPARTITION_COMPLEX, 84);
testRepartitionData(getSession(), sourceQuery, false, TEST_REPARTITION_COMPLEX, 84);
// with same column multiple times
testRepartitionData(getSession(), sourceQuery, true, TEST_REPARTITION_SAME_COLUMN_MULTIPLE_TIMES, 97);
testRepartitionData(getSession(), sourceQuery, false, TEST_REPARTITION_SAME_COLUMN_MULTIPLE_TIMES, 97);
}

@Test
@Override
public void testShowCreateTable()
{
assertThat((String) computeScalar("SHOW CREATE TABLE region"))
.isEqualTo("" +
"CREATE TABLE iceberg." + schemaName + ".region (\n" +
" regionkey bigint,\n" +
" name varchar,\n" +
" comment varchar\n" +
")\n" +
"WITH (\n" +
" format = '" + format + "',\n" +
" location = 's3://" + bucketName + "/" + schemaName + "/region'\n" +
")");
}

@Test
@Override
public void testRenameSchema()
{
// Overridden because error message is different from upper test method
String schemaName = getSession().getSchema().orElseThrow();
assertQueryFails(
format("ALTER SCHEMA %s RENAME TO %s", schemaName, schemaName + randomTableSuffix()),
"Hive metastore does not support renaming schemas");
}

@Test
@Override
public void testShowCreateSchema()
{
assertThat(computeActual("SHOW CREATE SCHEMA " + schemaName).getOnlyValue().toString())
.matches("CREATE SCHEMA iceberg." + schemaName + "\n" +
"AUTHORIZATION USER user\n" +
"WITH \\(\n" +
"\\s+location = 's3://" + bucketName + "/" + schemaName + "'\n" +
"\\)");
}
}
Original file line number Diff line number Diff line change
@@ -0,0 +1,62 @@
/*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package io.trino.plugin.iceberg;

import io.trino.Session;

public final class IcebergTestUtil
{
private IcebergTestUtil()
{
}

public static boolean orcSupportsIcebergFileStatistics(String typeName)
{
return !(typeName.equalsIgnoreCase("varbinary")) &&
!(typeName.equalsIgnoreCase("uuid"));
}

public static boolean orcSupportsRowGroupStatistics(String typeName)
{
return !typeName.equalsIgnoreCase("varbinary");
}

public static Session orcWithSmallRowGroups(Session session)
{
return Session.builder(session)
.setCatalogSessionProperty("iceberg", "orc_writer_max_stripe_rows", "10")
.build();
}

public static boolean parquetSupportsIcebergFileStatistics(String typeName)
{
return true;
}

public static boolean parquetSupportsRowGroupStatistics(String typeName)
{
return !(typeName.equalsIgnoreCase("varbinary") ||
typeName.equalsIgnoreCase("time(6)") ||
typeName.equalsIgnoreCase("timestamp(6) with time zone"));
}

public static Session parquetWithSmallRowGroups(Session session)
{
return Session.builder(session)
.setCatalogSessionProperty("iceberg", "parquet_writer_page_size", "100B")
.setCatalogSessionProperty("iceberg", "parquet_writer_block_size", "100B")
.setCatalogSessionProperty("iceberg", "parquet_writer_batch_size", "10")
.build();
}
}
Original file line number Diff line number Diff line change
@@ -0,0 +1,51 @@
/*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package io.trino.plugin.iceberg;

import io.trino.Session;
import org.testng.annotations.Test;

import static io.trino.plugin.iceberg.IcebergFileFormat.ORC;
import static io.trino.plugin.iceberg.IcebergTestUtil.orcSupportsIcebergFileStatistics;
import static io.trino.plugin.iceberg.IcebergTestUtil.orcSupportsRowGroupStatistics;
import static io.trino.plugin.iceberg.IcebergTestUtil.orcWithSmallRowGroups;

// Due to fact some tests running on MinIO are memory consuming, we want to prevent running them in parallel.
@Test(singleThreaded = true)
public class TestIcebergMinioOrcConnectorTest
extends BaseIcebergMinioConnectorTest
{
public TestIcebergMinioOrcConnectorTest()
{
super(ORC);
}

@Override
protected boolean supportsIcebergFileStatistics(String typeName)
{
return orcSupportsIcebergFileStatistics(typeName);
}

@Override
protected boolean supportsRowGroupStatistics(String typeName)
{
return orcSupportsRowGroupStatistics(typeName);
}

@Override
protected Session withSmallRowGroups(Session session)
{
return orcWithSmallRowGroups(session);
}
}
Loading

0 comments on commit 19e595f

Please sign in to comment.