Skip to content

add autosharding support #22512

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
Naireen wants to merge 1 commit into from
Closed

add autosharding support #22512

Show file tree
Hide file tree
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
6 changes: 2 additions & 4 deletions ...o/google-cloud-platform/src/main/java/org/apache/beam/sdk/io/gcp/bigquery/BigQueryIO.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2725,9 +2725,6 @@ public WriteResult expand(PCollection<T> input) {
if (input.isBounded() == IsBounded.BOUNDED) {
checkArgument(!getAutoSharding(), "Auto-sharding is only applicable to unbounded input.");
}
if (method == Write.Method.STORAGE_WRITE_API) {
checkArgument(!getAutoSharding(), "Auto sharding not yet available for Storage API writes");
}

if (getJsonTimePartitioning() != null) {
checkArgument(
Expand Down Expand Up @@ -3023,7 +3020,8 @@ private <DestinationT> WriteResult continueExpandTyped(
getStorageApiTriggeringFrequency(bqOptions),
getBigQueryServices(),
getStorageApiNumStreams(bqOptions),
method == Method.STORAGE_API_AT_LEAST_ONCE);
method == Method.STORAGE_API_AT_LEAST_ONCE,
getAutoSharding());
return input.apply("StorageApiLoads", storageApiLoads);
} else {
throw new RuntimeException("Unexpected write method " + method);
Expand Down
98 changes: 58 additions & 40 deletions ...gle-cloud-platform/src/main/java/org/apache/beam/sdk/io/gcp/bigquery/StorageApiLoads.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,6 +52,7 @@ public class StorageApiLoads<DestinationT, ElementT>
private final BigQueryServices bqServices;
private final int numShards;
private final boolean allowInconsistentWrites;
private final boolean allowAutosharding;

public StorageApiLoads(
Coder<DestinationT> destinationCoder,
Expand All @@ -61,7 +62,8 @@ public StorageApiLoads(
Duration triggeringFrequency,
BigQueryServices bqServices,
int numShards,
boolean allowInconsistentWrites) {
boolean allowInconsistentWrites,
boolean allowAutosharding) {
this.destinationCoder = destinationCoder;
this.dynamicDestinations = dynamicDestinations;
this.createDisposition = createDisposition;
Expand All @@ -70,6 +72,7 @@ public StorageApiLoads(
this.bqServices = bqServices;
this.numShards = numShards;
this.allowInconsistentWrites = allowInconsistentWrites;
this.allowAutosharding = allowAutosharding;
}

@Override
Expand Down Expand Up @@ -136,9 +139,6 @@ public WriteResult expandTriggered(
// Handle triggered, low-latency loads into BigQuery.
PCollection<KV<DestinationT, ElementT>> inputInGlobalWindow =
input.apply("rewindowIntoGlobal", Window.into(new GlobalWindows()));

// First shard all the records.
// TODO(reuvenlax): Add autosharding support so that users don't have to pick a shard count.
PCollectionTuple result =
inputInGlobalWindow
.apply(
Expand All @@ -154,43 +154,61 @@ public WriteResult expandTriggered(
successfulRowsTag,
BigQueryStorageApiInsertErrorCoder.of(),
successCoder));
PCollection<KV<ShardedKey<DestinationT>, StorageApiWritePayload>> shardedRecords =
result
.get(successfulRowsTag)
.apply(
"AddShard",
ParDo.of(
new DoFn<
KV<DestinationT, StorageApiWritePayload>,
KV<ShardedKey<DestinationT>, StorageApiWritePayload>>() {
int shardNumber;

@Setup
public void setup() {
shardNumber = ThreadLocalRandom.current().nextInt(numShards);
}

@ProcessElement
public void processElement(
@Element KV<DestinationT, StorageApiWritePayload> element,
OutputReceiver<KV<ShardedKey<DestinationT>, StorageApiWritePayload>> o) {
DestinationT destination = element.getKey();
ByteBuffer buffer = ByteBuffer.allocate(Integer.BYTES);
buffer.putInt(++shardNumber % numShards);
o.output(
KV.of(ShardedKey.of(destination, buffer.array()), element.getValue()));
}
}))
.setCoder(KvCoder.of(ShardedKey.Coder.of(destinationCoder), payloadCoder));

PCollection<KV<ShardedKey<DestinationT>, Iterable<StorageApiWritePayload>>> groupedRecords =
shardedRecords.apply(
"GroupIntoBatches",
GroupIntoBatches.<ShardedKey<DestinationT>, StorageApiWritePayload>ofByteSize(
MAX_BATCH_SIZE_BYTES,
(StorageApiWritePayload e) -> (long) e.getPayload().length)
.withMaxBufferingDuration(triggeringFrequency));

PCollection<KV<ShardedKey<DestinationT>, Iterable<StorageApiWritePayload>>> groupedRecords;

if (this.allowAutosharding) {
groupedRecords =
result
.get(successfulRowsTag)
.apply(
"GroupIntoBatches",
GroupIntoBatches.<DestinationT, StorageApiWritePayload>ofByteSize(
MAX_BATCH_SIZE_BYTES,
(StorageApiWritePayload e) -> (long) e.getPayload().length)
.withMaxBufferingDuration(triggeringFrequency)
.withShardedKey());

} else {
PCollection<KV<ShardedKey<DestinationT>, StorageApiWritePayload>> shardedRecords =
result
.get(successfulRowsTag)
.apply(
"AddShard",
ParDo.of(
new DoFn<
KV<DestinationT, StorageApiWritePayload>,
KV<ShardedKey<DestinationT>, StorageApiWritePayload>>() {
int shardNumber;

@Setup
public void setup() {
shardNumber = ThreadLocalRandom.current().nextInt(numShards);
}

@ProcessElement
public void processElement(
@Element KV<DestinationT, StorageApiWritePayload> element,
OutputReceiver<KV<ShardedKey<DestinationT>, StorageApiWritePayload>>
o) {
DestinationT destination = element.getKey();
ByteBuffer buffer = ByteBuffer.allocate(Integer.BYTES);
buffer.putInt(++shardNumber % numShards);
o.output(
KV.of(
ShardedKey.of(destination, buffer.array()), element.getValue()));
}
}))
.setCoder(KvCoder.of(ShardedKey.Coder.of(destinationCoder), payloadCoder));

groupedRecords =
shardedRecords.apply(
"GroupIntoBatches",
GroupIntoBatches.<ShardedKey<DestinationT>, StorageApiWritePayload>ofByteSize(
MAX_BATCH_SIZE_BYTES,
(StorageApiWritePayload e) -> (long) e.getPayload().length)
.withMaxBufferingDuration(triggeringFrequency));
}
groupedRecords.apply(
"StorageApiWriteSharded",
new StorageApiWritesShardedRecords<>(
Expand Down
22 changes: 15 additions & 7 deletions ...cloud-platform/src/test/java/org/apache/beam/sdk/io/gcp/bigquery/BigQueryIOWriteTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1190,19 +1190,21 @@ protected void writeString(org.apache.avro.Schema schema, Object datum, Encoder

@Test
public void testStreamingWrite() throws Exception {
streamingWrite(false);
autoShardingWrite(false, false);
}

@Test
public void testStreamingWriteWithAutoSharding() throws Exception {
if (useStorageApi) {
return;
}
streamingWrite(true);
autoShardingWrite(true, false);
}

private void streamingWrite(boolean autoSharding) throws Exception {
if (!useStreaming) {
@Test
public void testStorageApiWriteWithAutoSharding() throws Exception {
autoShardingWrite(true, true);
}

private void autoShardingWrite(boolean autoSharding, boolean storageAPIWrite) throws Exception {
if (!useStreaming && !storageAPIWrite) {
return;
}
BigQueryIO.Write<TableRow> write =
Expand All @@ -1220,6 +1222,12 @@ private void streamingWrite(boolean autoSharding) throws Exception {
if (autoSharding) {
write = write.withAutoSharding();
}
if (storageAPIWrite) {
write =
write
.withTriggeringFrequency(Duration.standardSeconds(5))
.withMethod(Method.STORAGE_WRITE_API);
}
p.apply(
Create.of(
new TableRow().set("name", "a").set("number", "1"),
Expand Down