Skip to content

DIFF PARTISH #2

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

Closed
deniskuzZ wants to merge 2 commits into from
Closed

DIFF PARTISH #2

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
9f3d9db
Iceberg part-level column stats calc support
deniskuzZ Feb 20, 2025
8106efb
added support for stats part evolution
deniskuzZ Mar 9, 2025
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
345 changes: 185 additions & 160 deletions ...g/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/HiveIcebergStorageHandler.java
View file
Open in desktop

Large diffs are not rendered by default.

36 changes: 0 additions & 36 deletions iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/HiveTableUtil.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,8 +19,6 @@

package org.apache.iceberg.mr.hive;

import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import java.io.IOException;
import java.io.ObjectInputStream;
import java.io.ObjectOutputStream;
Expand All @@ -34,14 +32,12 @@
import java.util.concurrent.Callable;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.stream.Collectors;
import java.util.stream.IntStream;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.LocatedFileStatus;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.fs.RemoteIterator;
import org.apache.hadoop.hive.conf.Constants;
import org.apache.hadoop.hive.conf.HiveConf;
import org.apache.hadoop.hive.metastore.Warehouse;
import org.apache.hadoop.hive.metastore.api.FieldSchema;
Expand All @@ -50,12 +46,8 @@
import org.apache.hadoop.hive.metastore.api.hive_metastoreConstants;
import org.apache.hadoop.hive.metastore.partition.spec.PartitionSpecProxy;
import org.apache.hadoop.hive.metastore.utils.FileUtils;
import org.apache.hadoop.hive.ql.exec.Utilities;
import org.apache.hadoop.hive.ql.io.IOConstants;
import org.apache.hadoop.hive.ql.parse.SemanticException;
import org.apache.hadoop.hive.serde.serdeConstants;
import org.apache.hadoop.hive.serde2.ColumnProjectionUtils;
import org.apache.hadoop.mapred.JobConf;
import org.apache.iceberg.AppendFiles;
import org.apache.iceberg.DataFile;
import org.apache.iceberg.DeleteFiles;
Expand Down Expand Up @@ -281,32 +273,4 @@ public static boolean isCtas(Properties properties) {
return Boolean.parseBoolean(properties.getProperty(hive_metastoreConstants.TABLE_IS_CTAS));
}

static Properties getSerializationProps() {
Properties props = new Properties();
props.put(serdeConstants.SERIALIZATION_FORMAT, "" + Utilities.tabCode);
props.put(serdeConstants.SERIALIZATION_NULL_FORMAT, "NULL");
return props;
}

static String getParseData(String parseData, String specId, ObjectMapper mapper, Integer currentSpecId)
throws JsonProcessingException {
Map<String, String> map = mapper.readValue(parseData, Map.class);
String partString =
map.entrySet().stream()
.filter(entry -> entry.getValue() != null)
.map(java.lang.Object::toString)
.collect(Collectors.joining("/"));
String currentSpecMarker = currentSpecId.toString().equals(specId) ? "current-" : "";
return String.format("%sspec-id=%s/%s", currentSpecMarker, specId, partString);
}

static JobConf getPartJobConf(Configuration confs, org.apache.hadoop.hive.ql.metadata.Table tbl) {
JobConf job = new JobConf(confs);
job.set(ColumnProjectionUtils.READ_COLUMN_NAMES_CONF_STR, Constants.ICEBERG_PARTITION_COLUMNS);
job.set(InputFormatConfig.TABLE_LOCATION, tbl.getPath().toString());
job.set(InputFormatConfig.TABLE_IDENTIFIER, tbl.getFullyQualifiedName() + ".partitions");
HiveConf.setVar(job, HiveConf.ConfVars.HIVE_FETCH_OUTPUT_SERDE, Constants.DELIMITED_JSON_SERDE);
HiveConf.setBoolVar(job, HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, false);
return job;
}
}
80 changes: 39 additions & 41 deletions iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/IcebergTableUtil.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,7 @@
import java.io.IOException;
import java.time.ZoneId;
import java.util.Collections;
import java.util.Comparator;
import java.util.List;
import java.util.Map;
import java.util.Optional;
Expand Down Expand Up @@ -473,50 +474,56 @@ public static List<FieldSchema> getPartitionKeys(Table table, int specId) {
public static List<PartitionField> getPartitionFields(Table table, boolean latestSpecOnly) {
return latestSpecOnly ? table.spec().fields() :
table.specs().values().stream()
.flatMap(spec -> spec.fields().stream()).distinct()
.flatMap(spec -> spec.fields().stream()
.filter(f -> !f.transform().isVoid()))
.distinct()
.collect(Collectors.toList());
}

/**
* Returns a Map of PartitionData as the keys and partition spec ids as the values
* @param icebergTable Iceberg table
* Returns a list of partition names satisfying the provided partition spec.
* @param table Iceberg table
* @param partSpecMap Partition Spec used as the criteria for filtering
* @param allowPartialSpec When true, must return partitions which match partSpecMap exactly, otherwise partially
* @param latestSpecOnly When true, returns partitions with the latest partition spec only, otherwise with any specs
* @return Map of PartitionData and partition spec found based on the specified constraints
* @param latestSpecOnly when True, returns partitions with the current spec only, else - any specs
* @return List of partition names
*/
public static Map<PartitionData, Integer> getPartitionInfo(Table icebergTable, Map<String, String> partSpecMap,
boolean allowPartialSpec, boolean latestSpecOnly) throws SemanticException, IOException {
public static List<String> getPartitionNames(Table table, Map<String, String> partSpecMap,
boolean latestSpecOnly) throws SemanticException {
Expression expression = IcebergTableUtil.generateExpressionFromPartitionSpec(
icebergTable, partSpecMap, latestSpecOnly);
table, partSpecMap, latestSpecOnly);
PartitionsTable partitionsTable = (PartitionsTable) MetadataTableUtils.createMetadataTableInstance(
icebergTable, MetadataTableType.PARTITIONS);
table, MetadataTableType.PARTITIONS);

Map<PartitionData, Integer> result = Maps.newLinkedHashMap();
try (CloseableIterable<FileScanTask> fileScanTasks = partitionsTable.newScan().planFiles()) {
fileScanTasks.forEach(task ->
CloseableIterable.filter(
CloseableIterable.transform(task.asDataTask().rows(), row -> {
StructProjection data = row.get(IcebergTableUtil.PART_IDX, StructProjection.class);
Integer specId = row.get(IcebergTableUtil.SPEC_IDX, Integer.class);
return Maps.immutableEntry(IcebergTableUtil.toPartitionData(data,
Partitioning.partitionType(icebergTable), icebergTable.specs().get(specId).partitionType()),
specId);
}), entry -> {
ResidualEvaluator resEval = ResidualEvaluator.of(icebergTable.specs().get(entry.getValue()),
expression, false);
return resEval.residualFor(entry.getKey()).isEquivalentTo(Expressions.alwaysTrue()) &&
(entry.getKey().size() == partSpecMap.size() || allowPartialSpec) &&
(entry.getValue() == icebergTable.spec().specId() || !latestSpecOnly);
}).forEach(entry -> result.put(entry.getKey(), entry.getValue())));
}
return FluentIterable.from(fileScanTasks)
.transformAndConcat(task -> task.asDataTask().rows())
.transform(row -> {
StructLike data = row.get(IcebergTableUtil.PART_IDX, StructProjection.class);
PartitionSpec spec = table.specs().get(row.get(IcebergTableUtil.SPEC_IDX, Integer.class));
return Maps.immutableEntry(
IcebergTableUtil.toPartitionData(
data, Partitioning.partitionType(table), spec.partitionType()),
spec);
}).filter(e -> {
ResidualEvaluator resEval = ResidualEvaluator.of(e.getValue(),
expression, false);
return e.getValue().isPartitioned() &&
resEval.residualFor(e.getKey()).isEquivalentTo(Expressions.alwaysTrue()) &&
(e.getValue().specId() == table.spec().specId() || !latestSpecOnly);

}).transform(e -> e.getValue().partitionToPath(e.getKey())).toSortedList(
Comparator.naturalOrder());

return result;
} catch (IOException e) {
throw new SemanticException(
String.format("Error while fetching the partitions due to: %s", e));
}
}

public static long getPartitionHash(Table icebergTable, String partitionPath) throws IOException {
PartitionsTable partitionsTable = (PartitionsTable) MetadataTableUtils
.createMetadataTableInstance(icebergTable, MetadataTableType.PARTITIONS);

try (CloseableIterable<FileScanTask> fileScanTasks = partitionsTable.newScan().planFiles()) {
return FluentIterable.from(fileScanTasks)
.transformAndConcat(task -> task.asDataTask().rows())
Expand All @@ -534,19 +541,10 @@ public static long getPartitionHash(Table icebergTable, String partitionPath) th
}
}

public static List<String> getPartitionNames(Table icebergTable, Map<String, String> partitionSpec,
boolean latestSpecOnly) throws SemanticException {
try {
return IcebergTableUtil
.getPartitionInfo(icebergTable, partitionSpec, true, latestSpecOnly).entrySet().stream()
.map(e -> {
PartitionData partitionData = e.getKey();
int specId = e.getValue();
return icebergTable.specs().get(specId).partitionToPath(partitionData);
}).collect(Collectors.toList());
} catch (IOException e) {
throw new SemanticException(String.format("Error while fetching the partitions due to: %s", e));
}
public static TransformSpec getTransformSpec(Table table, String transformName, int sourceId) {
TransformSpec spec = TransformSpec.fromString(transformName.toUpperCase(),
table.schema().findColumnName(sourceId));
return spec;
}

}
4 changes: 4 additions & 0 deletions ...-handler/src/test/java/org/apache/iceberg/mr/hive/HiveIcebergStorageHandlerTestUtils.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,6 +126,10 @@ static TestTables testTables(TestHiveShell shell, TestTables.TestTableType testT
return testTableType.instance(shell.metastore().hiveConf(), temp, catalogName);
}

static void init(TestHiveShell shell, TestTables testTables, TemporaryFolder temp) {
init(shell, testTables, temp, "tez");
}

static void init(TestHiveShell shell, TestTables testTables, TemporaryFolder temp, String engine) {
shell.getSession();

Expand Down
48 changes: 18 additions & 30 deletions ...ler/src/test/java/org/apache/iceberg/mr/hive/HiveIcebergStorageHandlerWithEngineBase.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,8 +33,8 @@
import org.apache.iceberg.SnapshotSummary;
import org.apache.iceberg.Table;
import org.apache.iceberg.data.Record;
import org.apache.iceberg.hive.HiveVersion;
import org.apache.iceberg.mr.TestHelper;
import org.apache.iceberg.mr.hive.TestTables.TestTableType;
import org.apache.iceberg.relocated.com.google.common.collect.ImmutableList;
import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap;
import org.apache.iceberg.relocated.com.google.common.collect.Lists;
Expand All @@ -53,6 +53,8 @@
import org.junit.runner.RunWith;
import org.junit.runners.Parameterized;

import static org.apache.iceberg.mr.hive.TestTables.ALL_TABLE_TYPES;
import static org.apache.iceberg.mr.hive.TestTables.TestTableType.HIVE_CATALOG;
import static org.apache.iceberg.types.Types.NestedField.optional;
import static org.apache.iceberg.types.Types.NestedField.required;
import static org.junit.Assume.assumeTrue;
Expand All @@ -62,8 +64,6 @@
@RunWith(Parameterized.class)
public abstract class HiveIcebergStorageHandlerWithEngineBase {

protected static final String[] EXECUTION_ENGINES = new String[] {"tez"};

public static final String RETRY_STRATEGIES =
"overlay,reoptimize,reexecute_lost_am,dagsubmit,recompile_without_cbo,write_conflict";

Expand Down Expand Up @@ -108,34 +108,25 @@ public abstract class HiveIcebergStorageHandlerWithEngineBase {
StatsSetupConst.TOTAL_SIZE, SnapshotSummary.TOTAL_FILE_SIZE_PROP
);

@Parameters(name = "fileFormat={0}, engine={1}, catalog={2}, isVectorized={3}, formatVersion={4}")
@Parameters(name = "fileFormat={0}, catalog={1}, isVectorized={2}, formatVersion={3}")
public static Collection<Object[]> parameters() {
Collection<Object[]> testParams = Lists.newArrayList();
String javaVersion = System.getProperty("java.specification.version");

// Run tests with every FileFormat for a single Catalog (HiveCatalog)
for (FileFormat fileFormat : HiveIcebergStorageHandlerTestUtils.FILE_FORMATS) {
for (String engine : EXECUTION_ENGINES) {
IntStream.of(2, 1).forEach(formatVersion -> {
// include Tez tests only for Java 8
if (javaVersion.equals("1.8")) {
testParams.add(new Object[]{fileFormat, engine, TestTables.TestTableType.HIVE_CATALOG, false,
formatVersion});
// test for vectorization=ON in case of ORC and PARQUET format with Tez engine
if (fileFormat != FileFormat.METADATA && "tez".equals(engine) && HiveVersion.min(HiveVersion.HIVE_3)) {
testParams.add(new Object[]{fileFormat, engine, TestTables.TestTableType.HIVE_CATALOG, true,
formatVersion});
}
}
});
}
IntStream.of(2, 1).forEach(formatVersion -> {
testParams.add(new Object[]{fileFormat, HIVE_CATALOG, false, formatVersion});
// test for vectorization=ON in case of ORC and PARQUET format
if (fileFormat != FileFormat.METADATA) {
testParams.add(new Object[]{fileFormat, HIVE_CATALOG, true, formatVersion});
}
});
}

// Run tests for every Catalog for a single FileFormat (PARQUET) and execution engine (tez)
// skip HiveCatalog tests as they are added before
for (TestTables.TestTableType testTableType : TestTables.ALL_TABLE_TYPES) {
if (!TestTables.TestTableType.HIVE_CATALOG.equals(testTableType)) {
testParams.add(new Object[]{FileFormat.PARQUET, "tez", testTableType, false, 1});
// Run tests for every Catalog for a single FileFormat (PARQUET), skip HiveCatalog tests as they are added before
for (TestTableType testTableType : ALL_TABLE_TYPES) {
if (testTableType != HIVE_CATALOG) {
testParams.add(new Object[]{FileFormat.PARQUET, testTableType, false, 1});
}
}

Expand All @@ -150,15 +141,12 @@ public static Collection<Object[]> parameters() {
public FileFormat fileFormat;

@Parameter(1)
public String executionEngine;
public TestTableType testTableType;

@Parameter(2)
public TestTables.TestTableType testTableType;

@Parameter(3)
public boolean isVectorized;

@Parameter(4)
@Parameter(3)
public Integer formatVersion;

@Rule
Expand All @@ -181,7 +169,7 @@ public static void afterClass() throws Exception {
public void before() throws IOException {
validateTestParams();
testTables = HiveIcebergStorageHandlerTestUtils.testTables(shell, testTableType, temp);
HiveIcebergStorageHandlerTestUtils.init(shell, testTables, temp, executionEngine);
HiveIcebergStorageHandlerTestUtils.init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
// Fetch task conversion might kick in for certain queries preventing vectorization code path to be used, so
// we turn it off explicitly to achieve better coverage.
Expand Down
10 changes: 5 additions & 5 deletions ...rg/iceberg-handler/src/test/java/org/apache/iceberg/mr/hive/TestConflictingDataFiles.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,7 +98,7 @@ public void testSingleFilterUpdate() {
try {
Tasks.range(2).executeWith(Executors.newFixedThreadPool(2)).run(i -> {
TestUtilPhaser.getInstance().getPhaser().register();
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -141,7 +141,7 @@ public void testMultiFiltersUpdate() {
try {
Tasks.range(2).executeWith(Executors.newFixedThreadPool(2)).run(i -> {
TestUtilPhaser.getInstance().getPhaser().register();
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -188,7 +188,7 @@ public void testDeleteFilters() {
try {
Tasks.range(3).executeWith(Executors.newFixedThreadPool(3)).run(i -> {
TestUtilPhaser.getInstance().getPhaser().register();
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -230,7 +230,7 @@ public void testConflictingUpdates() {
try {
Tasks.range(2).executeWith(Executors.newFixedThreadPool(2)).run(i -> {
TestUtilPhaser.getInstance().getPhaser().register();
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -299,7 +299,7 @@ public void testConcurrentInsertAndInsertOverwrite() {
throw new RuntimeException(e);
}
}
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");

Expand Down
8 changes: 4 additions & 4 deletions iceberg/iceberg-handler/src/test/java/org/apache/iceberg/mr/hive/TestHiveIcebergCRUD.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -616,7 +616,7 @@ public void testConcurrent2Deletes() {
Tasks.range(2)
.executeWith(Executors.newFixedThreadPool(2))
.run(i -> {
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -647,7 +647,7 @@ public void testConcurrent2Updates() {
Tasks.range(2)
.executeWith(Executors.newFixedThreadPool(2))
.run(i -> {
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -682,7 +682,7 @@ public void testConcurrentUpdateAndDelete() {
Tasks.range(2)
.executeWith(Executors.newFixedThreadPool(2))
.run(i -> {
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down Expand Up @@ -717,7 +717,7 @@ public void testConcurrent2MergeInserts() {
Tasks.range(2)
.executeWith(Executors.newFixedThreadPool(2))
.run(i -> {
init(shell, testTables, temp, executionEngine);
init(shell, testTables, temp);
HiveConf.setBoolVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_VECTORIZATION_ENABLED, isVectorized);
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_FETCH_TASK_CONVERSION, "none");
HiveConf.setVar(shell.getHiveConf(), HiveConf.ConfVars.HIVE_QUERY_REEXECUTION_STRATEGIES,
Expand Down
Loading
Loading