Skip to content
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

Jump to bottom

[BEAM-12482] Respect schemaUpdateOptions during BigQuery load with temporary tables #16303

Closed
wants to merge 2 commits into from
Closed

[BEAM-12482] Respect schemaUpdateOptions during BigQuery load with temporary tables #16303

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
9eb57c1
Added a zero row job to bigquery writeTable
MiguelAnzoWizeline Dec 21, 2021
ebc47fd
added unit test
MiguelAnzoWizeline Dec 22, 2021
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
141 changes: 141 additions & 0 deletions .../google-cloud-platform/src/main/java/org/apache/beam/sdk/io/gcp/bigquery/WriteTables.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,13 +23,15 @@
import com.google.api.services.bigquery.model.EncryptionConfiguration;
import com.google.api.services.bigquery.model.JobConfigurationLoad;
import com.google.api.services.bigquery.model.JobReference;
import com.google.api.services.bigquery.model.Table;
import com.google.api.services.bigquery.model.TableReference;
import com.google.api.services.bigquery.model.TableSchema;
import com.google.api.services.bigquery.model.TimePartitioning;
import com.google.auto.value.AutoValue;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.util.Collections;
import java.util.List;
import java.util.Map;
import java.util.Set;
Expand Down Expand Up @@ -269,6 +271,19 @@ public void processElement(
createDisposition = CreateDisposition.CREATE_IF_NEEDED;
}

BigQueryHelpers.PendingJob schemaJob =
startZeroLoadJob(
getJobService(c.getPipelineOptions().as(BigQueryOptions.class)),
getDatasetService(c.getPipelineOptions().as(BigQueryOptions.class)),
jobIdPrefix,
tableReference,
tableDestination.getTimePartitioning(),
tableDestination.getClustering(),
tableSchema,
writeDisposition,
createDisposition,
schemaUpdateOptions);

BigQueryHelpers.PendingJob retryJob =
startLoad(
getJobService(c.getPipelineOptions().as(BigQueryOptions.class)),
Expand All @@ -282,6 +297,16 @@ public void processElement(
writeDisposition,
createDisposition,
schemaUpdateOptions);
if (schemaJob != null) {
pendingJobs.add(
new PendingJobData(
window,
schemaJob,
partitionFiles,
tableDestination,
tableReference,
element.getValue().isFirstPane()));
}
pendingJobs.add(
new PendingJobData(
window,
Expand Down Expand Up @@ -549,6 +574,122 @@ private PendingJob startLoad(
return retryJob;
}

private PendingJob startZeroLoadJob(
JobService jobService,
DatasetService datasetService,
String jobIdPrefix,
TableReference ref,
TimePartitioning timePartitioning,
Clustering clustering,
@Nullable TableSchema schema,
WriteDisposition writeDisposition,
CreateDisposition createDisposition,
Set<SchemaUpdateOption> schemaUpdateOptions) {
JobConfigurationLoad loadConfig =
new JobConfigurationLoad()
.setDestinationTable(ref)
.setSchema(schema)
.setSourceUris(Collections.EMPTY_LIST)
.setWriteDisposition(writeDisposition.name())
.setCreateDisposition(createDisposition.name())
.setSourceFormat(sourceFormat)
.setIgnoreUnknownValues(ignoreUnknownValues)
.setUseAvroLogicalTypes(useAvroLogicalTypes);
if (schemaUpdateOptions != null) {
List<String> options =
schemaUpdateOptions.stream()
.map(Enum<SchemaUpdateOption>::name)
.collect(Collectors.toList());
loadConfig.setSchemaUpdateOptions(options);
}
if (!loadConfig.getWriteDisposition().equals(WriteDisposition.WRITE_TRUNCATE.toString())
|| !loadConfig.getWriteDisposition().equals(WriteDisposition.WRITE_APPEND.toString())) {
return null;
}
Table destinationTable = null;
try {
destinationTable = datasetService.getTable(ref);
if (destinationTable == null) {
return null; // no need to update schema ahead if table does not exists
}
} catch (IOException | InterruptedException e) {
LOG.warn("Failed to get table {} with {}", ref, e.toString());
throw new RuntimeException(e);
}
if (destinationTable.getSchema().equals(schema)) {
return null; // no need to update schema ahead if schema is already the same
}
if (timePartitioning != null) {
loadConfig.setTimePartitioning(timePartitioning);
// only set clustering if timePartitioning is set
if (clustering != null) {
loadConfig.setClustering(clustering);
}
}
if (kmsKey != null) {
loadConfig.setDestinationEncryptionConfiguration(
new EncryptionConfiguration().setKmsKeyName(kmsKey));
}
String projectId =
loadJobProjectId == null || loadJobProjectId.get() == null
? ref.getProjectId()
: loadJobProjectId.get();
String bqLocation =
BigQueryHelpers.getDatasetLocation(datasetService, ref.getProjectId(), ref.getDatasetId());

PendingJob retryJob =
new PendingJob(
// Function to load the data.
jobId -> {
JobReference jobRef =
new JobReference()
.setProjectId(projectId)
.setJobId(jobId.getJobId())
.setLocation(bqLocation);
LOG.info(
"Loading zero rows using job {}, job id {} iteration {}",
ref,
jobRef,
jobId.getRetryIndex());
try {
jobService.startLoadJob(jobRef, loadConfig);
} catch (IOException | InterruptedException e) {
LOG.warn("Load job {} failed with {}", jobRef, e.toString());
throw new RuntimeException(e);
}
return null;
},
// Function to poll the result of a load job.
jobId -> {
JobReference jobRef =
new JobReference()
.setProjectId(projectId)
.setJobId(jobId.getJobId())
.setLocation(bqLocation);
try {
return jobService.pollJob(jobRef, BatchLoads.LOAD_JOB_POLL_MAX_RETRIES);
} catch (InterruptedException e) {
throw new RuntimeException(e);
}
},
// Function to lookup a job.
jobId -> {
JobReference jobRef =
new JobReference()
.setProjectId(projectId)
.setJobId(jobId.getJobId())
.setLocation(bqLocation);
try {
return jobService.getJob(jobRef);
} catch (InterruptedException | IOException e) {
throw new RuntimeException(e);
}
},
maxRetryJobs,
jobIdPrefix);
return retryJob;
}

static void removeTemporaryFiles(Iterable<String> files) throws IOException {
ImmutableList.Builder<ResourceId> fileResources = ImmutableList.builder();
for (String file : files) {
Expand Down
60 changes: 60 additions & 0 deletions ...form/src/test/java/org/apache/beam/sdk/io/gcp/bigquery/BigQuerySchemaUpdateOptionsIT.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -219,4 +219,64 @@ public void testAllowFieldRelaxation() throws Exception {
List<List<String>> expectedResult = Arrays.asList(Arrays.asList(value));
runWriteTest(schemaUpdateOptions, tableName, newSchema, rowToInsert, testQuery, expectedResult);
}

@Test
public void runWriteTestTempTables() throws Exception {
String tableName = makeTestTable();

Set<SchemaUpdateOption> schemaUpdateOptions =
EnumSet.of(BigQueryIO.Write.SchemaUpdateOption.ALLOW_FIELD_ADDITION);

TableSchema schema =
new TableSchema()
.setFields(
ImmutableList.of(
new TableFieldSchema().setName("new_field").setType("STRING"),
new TableFieldSchema().setName("optional_field").setType("STRING"),
new TableFieldSchema()
.setName("required_field")
.setType("STRING")
.setMode("REQUIRED")));

String[] values = {"meow", "bark"};
TableRow rowToInsert =
new TableRow().set("new_field", values[0]).set("required_field", values[1]);
Copy link
Member

Choose a reason for hiding this comment

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

is a single row enough to cause two loads to happen? Can we please add another Row so that we can force a two-stage load?


String testQuery =
String.format(
"SELECT new_field, required_field FROM [%s.%s];", BIG_QUERY_DATASET_ID, tableName);

List<List<String>> expectedResult = Arrays.asList(Arrays.asList(values));
Options options = TestPipeline.testingPipelineOptions().as(Options.class);
options.setTempLocation(options.getTempRoot() + "/bq_it_temp");

Pipeline p = Pipeline.create(options);
Create.Values<TableRow> input = Create.<TableRow>of(rowToInsert);

Write<TableRow> writer =
BigQueryIO.writeTableRows()
.to(String.format("%s:%s.%s", options.getProject(), BIG_QUERY_DATASET_ID, tableName))
.withSchema(schema)
.withCreateDisposition(BigQueryIO.Write.CreateDisposition.CREATE_IF_NEEDED)
.withWriteDisposition(BigQueryIO.Write.WriteDisposition.WRITE_APPEND)
.withSchemaUpdateOptions(schemaUpdateOptions)
.withMaxBytesPerPartition(1)
.withMaxFilesPerPartition(1);

p.apply(input).apply(writer);
p.run().waitUntilFinish();

QueryResponse response = BQ_CLIENT.queryWithRetries(testQuery, project);

List<List<String>> result =
response.getRows().stream()
.map(
row ->
row.getF().stream()
.map(cell -> cell.getV().toString())
.collect(Collectors.toList()))
.collect(Collectors.toList());

assertEquals(expectedResult, result);
}
}