Skip to content

Fix GlueHiveMetastore#batchGetPartition to handle unprocessedKeys #6155

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
rentaow wants to merge 1 commit into from
Closed

Fix GlueHiveMetastore#batchGetPartition to handle unprocessedKeys #6155

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
45 changes: 30 additions & 15 deletions presto-hive/src/main/java/io/prestosql/plugin/hive/metastore/glue/GlueHiveMetastore.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -870,24 +870,39 @@ private Map<String, Optional<Partition>> getPartitionsByNames(Table table, List<
private List<Partition> batchGetPartition(Table table, List<String> partitionNames)
{
try {
List<Future<BatchGetPartitionResult>> batchGetPartitionFutures = new ArrayList<>();

for (List<String> partitionNamesBatch : Lists.partition(partitionNames, BATCH_GET_PARTITION_MAX_PAGE_SIZE)) {
List<PartitionValueList> partitionValuesBatch = mappedCopy(partitionNamesBatch, partitionName -> new PartitionValueList().withValues(toPartitionValues(partitionName)));
batchGetPartitionFutures.add(glueClient.batchGetPartitionAsync(new BatchGetPartitionRequest()
.withCatalogId(catalogId)
.withDatabaseName(table.getDatabaseName())
.withTableName(table.getTableName())
.withPartitionsToGet(partitionValuesBatch)));
}
List<PartitionValueList> partitionValueLists = mappedCopy(partitionNames, partitionName -> new PartitionValueList().withValues(toPartitionValues(partitionName)));
ImmutableList.Builder<Partition> resultsBuilder = ImmutableList.builderWithExpectedSize(partitionNames.size());

// Reuse immutable field instances opportunistically between partitions
GluePartitionConverter converter = new GluePartitionConverter(table);
ImmutableList.Builder<Partition> resultsBuilder = ImmutableList.builderWithExpectedSize(partitionNames.size());
for (Future<BatchGetPartitionResult> future : batchGetPartitionFutures) {
future.get().getPartitions().stream()
.map(converter)
.forEach(resultsBuilder::add);

while (!partitionValueLists.isEmpty()) {
List<List<PartitionValueList>> batchedPartitionValueLists = Lists.partition(partitionValueLists, BATCH_GET_PARTITION_MAX_PAGE_SIZE);
List<Future<BatchGetPartitionResult>> batchGetPartitionFutures = new ArrayList<>();
for (List<PartitionValueList> partitions : batchedPartitionValueLists) {
batchGetPartitionFutures.add(glueClient.batchGetPartitionAsync(new BatchGetPartitionRequest()
.withCatalogId(catalogId)
.withDatabaseName(table.getDatabaseName())
.withTableName(table.getTableName())
.withPartitionsToGet(partitions)));
}

List<PartitionValueList> unprocessedPartitions = new ArrayList<>();
for (Future<BatchGetPartitionResult> future : batchGetPartitionFutures) {
BatchGetPartitionResult batchGetPartitionResult = future.get();

// In the unlikely scenario where batchGetPartition call cannot make progress on retrieving partitions,
// we throw exception on this metastore operation to avoid getting stuck in this loop.
if (batchGetPartitionResult.getPartitions().isEmpty() && !batchGetPartitionResult.getUnprocessedKeys().isEmpty()) {
throw new PrestoException(HIVE_METASTORE_ERROR, "Unable to retrieve partitions: " + batchGetPartitionResult.getUnprocessedKeys());
}

batchGetPartitionResult.getPartitions().stream()
.map(converter)
.forEach(resultsBuilder::add);
unprocessedPartitions.addAll(batchGetPartitionResult.getUnprocessedKeys());
}
partitionValueLists = unprocessedPartitions;
}

return resultsBuilder.build();
Expand Down