Skip to content

Add option to use LakeFormation in S3Glue data source. #2624

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
vamsimanohar merged 2 commits into from
Apr 29, 2024
Merged

Add option to use LakeFormation in S3Glue data source. #2624

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5434526
Add option to use LakeFormation in S3Glue data source.
asuresh8 Apr 17, 2024
08ccd67
Update s3glue_connector.rst
vamsimanohar Apr 25, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions datasources/src/main/java/org/opensearch/sql/datasources/glue/GlueDataSourceFactory.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ public class GlueDataSourceFactory implements DataSourceFactory {
"glue.indexstore.opensearch.auth.password";
public static final String GLUE_INDEX_STORE_OPENSEARCH_REGION =
"glue.indexstore.opensearch.region";
public static final String GLUE_LAKEFORMATION_ENABLED = "glue.lakeformation.enabled";

@Override
public DataSourceType getDataSourceType() {
Expand Down
8 changes: 5 additions & 3 deletions docs/user/ppl/admin/connectors/s3glue_connector.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@ s3Glue connector provides a way to query s3 files using glue as metadata store a
This page covers s3Glue datasource configuration and also how to query and s3Glue datasource.

Required resources for s3 Glue Connector
===================================
========================================
* ``EMRServerless Spark Execution Engine Config Setting``: Since we execute s3Glue queries on top of spark execution engine, we require this configuration.
More details: `ExecutionEngine Config <../../../interfaces/asyncqueryinterface.rst#id2>`_
* ``S3``: This is where the data lies.
Expand All @@ -42,6 +42,7 @@ Glue Connector Properties.
* Basic Auth required ``glue.indexstore.opensearch.auth.username`` and ``glue.indexstore.opensearch.auth.password``
* AWSSigV4 Auth requires ``glue.indexstore.opensearch.auth.region`` and ``glue.auth.role_arn``
* ``glue.indexstore.opensearch.region`` [Required for awssigv4 auth]
* ``glue.lakeformation.enabled`` determines whether to enable lakeformation for queries. Default value is ``"false"`` if not specified

Sample Glue dataSource configuration
========================================
Expand All @@ -56,7 +57,7 @@ Glue datasource configuration::
"glue.auth.role_arn": "role_arn",
"glue.indexstore.opensearch.uri": "http://localhost:9200",
"glue.indexstore.opensearch.auth" :"basicauth",
"glue.indexstore.opensearch.auth.username" :"username"
"glue.indexstore.opensearch.auth.username" :"username",
"glue.indexstore.opensearch.auth.password" :"password"
},
"resultIndex": "query_execution_result"
Expand All @@ -71,6 +72,7 @@ Glue datasource configuration::
"glue.indexstore.opensearch.uri": "http://adsasdf.amazonopensearch.com:9200",
"glue.indexstore.opensearch.auth" :"awssigv4",
"glue.indexstore.opensearch.auth.region" :"awssigv4",
"glue.lakeformation.enabled": "true"
},
"resultIndex": "query_execution_result"
}]
Expand All @@ -86,4 +88,4 @@ Sample Queries

These queries would work only top of async queries. Documentation: `Async Query APIs <../../../interfaces/asyncqueryinterface.rst>`_

Documentation for Index Queries: https://github.com/opensearch-project/opensearch-spark/blob/main/docs/index.md
Documentation for Index Queries: https://github.com/opensearch-project/opensearch-spark/blob/main/docs/index.md
7 changes: 7 additions & 0 deletions spark/src/main/java/org/opensearch/sql/spark/asyncquery/model/SparkSubmitParameters.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,7 @@
import static org.opensearch.sql.datasources.glue.GlueDataSourceFactory.GLUE_INDEX_STORE_OPENSEARCH_AUTH_USERNAME;
import static org.opensearch.sql.datasources.glue.GlueDataSourceFactory.GLUE_INDEX_STORE_OPENSEARCH_REGION;
import static org.opensearch.sql.datasources.glue.GlueDataSourceFactory.GLUE_INDEX_STORE_OPENSEARCH_URI;
import static org.opensearch.sql.datasources.glue.GlueDataSourceFactory.GLUE_LAKEFORMATION_ENABLED;
import static org.opensearch.sql.datasources.glue.GlueDataSourceFactory.GLUE_ROLE_ARN;
import static org.opensearch.sql.spark.data.constants.SparkConstants.*;
import static org.opensearch.sql.spark.execution.statestore.StateStore.DATASOURCE_TO_REQUEST_INDEX;
Expand All @@ -21,6 +22,7 @@
import java.util.function.Supplier;
import lombok.AllArgsConstructor;
import lombok.RequiredArgsConstructor;
import org.apache.commons.lang3.BooleanUtils;
import org.apache.commons.text.StringEscapeUtils;
import org.opensearch.sql.datasource.model.DataSourceMetadata;
import org.opensearch.sql.datasource.model.DataSourceType;
Expand Down Expand Up @@ -117,6 +119,11 @@ public Builder dataSource(DataSourceMetadata metadata) {
config.put("spark.sql.catalog." + metadata.getName(), FLINT_DELEGATE_CATALOG);
config.put(FLINT_DATA_SOURCE_KEY, metadata.getName());

final boolean lakeFormationEnabled =
BooleanUtils.toBoolean(metadata.getProperties().get(GLUE_LAKEFORMATION_ENABLED));
config.put(EMR_LAKEFORMATION_OPTION, Boolean.toString(lakeFormationEnabled));
config.put(FLINT_ACCELERATE_USING_COVERING_INDEX, Boolean.toString(!lakeFormationEnabled));

setFlintIndexStoreHost(
parseUri(
metadata.getProperties().get(GLUE_INDEX_STORE_OPENSEARCH_URI), metadata.getName()));
Expand Down
5 changes: 5 additions & 0 deletions spark/src/main/java/org/opensearch/sql/spark/data/constants/SparkConstants.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,4 +105,9 @@ public class SparkConstants {
public static final String SPARK_CATALOG_CATALOG_IMPL =
"spark.sql.catalog.spark_catalog.catalog-impl";
public static final String ICEBERG_GLUE_CATALOG = "org.apache.iceberg.aws.glue.GlueCatalog";

public static final String EMR_LAKEFORMATION_OPTION =
"spark.emr-serverless.lakeformation.enabled";
public static final String FLINT_ACCELERATE_USING_COVERING_INDEX =
"spark.flint.optimizer.covering.enabled";
}
81 changes: 77 additions & 4 deletions spark/src/test/java/org/opensearch/sql/spark/dispatcher/SparkQueryDispatcherTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -168,6 +168,52 @@ void testDispatchSelectQuery() {
verifyNoInteractions(flintIndexMetadataService);
}

@Test
void testDispatchSelectQueryWithLakeFormation() {
HashMap<String, String> tags = new HashMap<>();
tags.put(DATASOURCE_TAG_KEY, "my_glue");
tags.put(CLUSTER_NAME_TAG_KEY, TEST_CLUSTER_NAME);
tags.put(JOB_TYPE_TAG_KEY, JobType.BATCH.getText());
String query = "select * from my_glue.default.http_logs";
String sparkSubmitParameters =
constructExpectedSparkSubmitParameterString(
"sigv4",
new HashMap<>() {
{
put(FLINT_INDEX_STORE_AWSREGION_KEY, "eu-west-1");
}
},
query,
true);
StartJobRequest expected =
new StartJobRequest(
"TEST_CLUSTER:batch",
EMRS_APPLICATION_ID,
EMRS_EXECUTION_ROLE,
sparkSubmitParameters,
tags,
false,
"query_execution_result_my_glue");
when(emrServerlessClient.startJobRun(expected)).thenReturn(EMR_JOB_ID);
DataSourceMetadata dataSourceMetadata = constructMyGlueDataSourceMetadataWithLakeFormation();
when(dataSourceService.verifyDataSourceAccessAndGetRawMetadata("my_glue"))
.thenReturn(dataSourceMetadata);

DispatchQueryResponse dispatchQueryResponse =
sparkQueryDispatcher.dispatch(
new DispatchQueryRequest(
EMRS_APPLICATION_ID,
query,
"my_glue",
LangType.SQL,
EMRS_EXECUTION_ROLE,
TEST_CLUSTER_NAME));
verify(emrServerlessClient, times(1)).startJobRun(startJobRequestArgumentCaptor.capture());
Assertions.assertEquals(expected, startJobRequestArgumentCaptor.getValue());
Assertions.assertEquals(EMR_JOB_ID, dispatchQueryResponse.getJobId());
verifyNoInteractions(flintIndexMetadataService);
}

@Test
void testDispatchSelectQueryWithBasicAuthIndexStoreDatasource() {
HashMap<String, String> tags = new HashMap<>();
Expand Down Expand Up @@ -936,13 +982,17 @@ void testDispatchQueryWithExtraSparkSubmitParameters() {

private String constructExpectedSparkSubmitParameterString(
String auth, Map<String, String> authParams, String query) {
return constructExpectedSparkSubmitParameterString(auth, authParams, query, false);
}

private String constructExpectedSparkSubmitParameterString(
String auth, Map<String, String> authParams, String query, boolean lakeFormationEnabled) {
StringBuilder authParamConfigBuilder = new StringBuilder();
for (String key : authParams.keySet()) {
authParamConfigBuilder.append(" --conf ");
authParamConfigBuilder.append(" --conf ");
authParamConfigBuilder.append(key);
authParamConfigBuilder.append("=");
authParamConfigBuilder.append(authParams.get(key));
authParamConfigBuilder.append(" ");
}
query = "\"" + query + "\"";
return " --class org.apache.spark.sql.FlintJob --conf"
Expand Down Expand Up @@ -978,9 +1028,13 @@ private String constructExpectedSparkSubmitParameterString(
+ " --conf"
+ " spark.hive.metastore.glue.role.arn=arn:aws:iam::924196221507:role/FlintOpensearchServiceRole"
+ " --conf spark.sql.catalog.my_glue=org.opensearch.sql.FlintDelegatingSessionCatalog "
+ " --conf spark.flint.datasource.name=my_glue "
+ " --conf spark.flint.datasource.name=my_glue --conf"
+ " spark.emr-serverless.lakeformation.enabled="
+ Boolean.toString(lakeFormationEnabled)
+ " --conf spark.flint.optimizer.covering.enabled="
+ Boolean.toString(!lakeFormationEnabled)
+ authParamConfigBuilder
+ " --conf spark.flint.job.query="
+ " --conf spark.flint.job.query="
+ query
+ " ";
}
Expand Down Expand Up @@ -1056,6 +1110,25 @@ private DataSourceMetadata constructMyGlueDataSourceMetadataWithBadURISyntax() {
.build();
}

private DataSourceMetadata constructMyGlueDataSourceMetadataWithLakeFormation() {

Map<String, String> properties = new HashMap<>();
properties.put("glue.auth.type", "iam_role");
properties.put(
"glue.auth.role_arn", "arn:aws:iam::924196221507:role/FlintOpensearchServiceRole");
properties.put(
"glue.indexstore.opensearch.uri",
"https://search-flint-dp-benchmark-cf5crj5mj2kfzvgwdeynkxnefy.eu-west-1.es.amazonaws.com");
Comment on lines +1117 to +1121
Copy link
Collaborator

@dai-chen dai-chen Apr 23, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I assume these account and domain info are internal? Can we replace with fake one for test?

Copy link
Contributor Author

@asuresh8 asuresh8 Apr 24, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It looks like these values are used for all tests. I can update as part of a follow-up PR.

properties.put("glue.indexstore.opensearch.auth", "awssigv4");
properties.put("glue.indexstore.opensearch.region", "eu-west-1");
properties.put("glue.lakeformation.enabled", "true");
return new DataSourceMetadata.Builder()
.setName("my_glue")
.setConnector(DataSourceType.S3GLUE)
.setProperties(properties)
.build();
}

private DataSourceMetadata constructPrometheusDataSourceType() {
return new DataSourceMetadata.Builder()
.setName("my_prometheus")
Expand Down