Cloud Spanner to Cloud Storage Text 템플릿은 Cloud Spanner 테이블에서 데이터를 읽고 Cloud Storage에 CSV 텍스트 파일로 쓰는 일괄 파이프라인입니다.
파이프라인 요구사항
파이프라인을 실행하기 전에 입력 Spanner 테이블이 있어야 합니다.
템플릿 매개변수
매개변수
설명
spannerProjectId
데이터를 읽을 Cloud Spanner 데이터베이스의 Google Cloud 프로젝트 ID입니다.
spannerDatabaseId
요청된 테이블의 데이터베이스 ID입니다.
spannerInstanceId
요청된 테이블의 인스턴스 ID입니다.
spannerTable
데이터를 읽을 테이블입니다.
textWritePrefix
출력 텍스트 파일이 기록되는 디렉터리입니다. 마지막에 /를 추가합니다. 예를 들면 gs://mybucket/somefolder/입니다.
spannerSnapshotTime
(선택사항) 읽으려는 Cloud Spanner 데이터베이스의 버전에 해당하는 타임스탬프입니다. 타임스탬프는 RFC 3339 UTC 'Zulu' 형식에 따라 지정되어야 합니다.
예를 들면 1990-12-31T23:59:60Z입니다. 타임스탬프는 과거여야 하며 최대 타임스탬프 비활성이 적용됩니다.
/*
* Copyright (C) 2018 Google LLC
*
* Licensed under the Apache License, Version 2.0 (the "License"); you may not
* use this file except in compliance with the License. You may obtain a copy of
* the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations under
* the License.
*/
package com.google.cloud.teleport.templates;
import static com.google.cloud.teleport.util.ValueProviderUtils.eitherOrValueProvider;
import com.google.cloud.spanner.Options.RpcPriority;
import com.google.cloud.teleport.metadata.Template;
import com.google.cloud.teleport.metadata.TemplateCategory;
import com.google.cloud.teleport.metadata.TemplateParameter;
import com.google.cloud.teleport.templates.SpannerToText.SpannerToTextOptions;
import com.google.cloud.teleport.templates.common.SpannerConverters;
import com.google.cloud.teleport.templates.common.SpannerConverters.CreateTransactionFnWithTimestamp;
import com.google.cloud.teleport.templates.common.SpannerConverters.SpannerReadOptions;
import com.google.cloud.teleport.templates.common.TextConverters.FilesystemWriteOptions;
import org.apache.beam.sdk.Pipeline;
import org.apache.beam.sdk.io.FileSystems;
import org.apache.beam.sdk.io.TextIO;
import org.apache.beam.sdk.io.fs.ResourceId;
import org.apache.beam.sdk.io.gcp.spanner.LocalSpannerIO;
import org.apache.beam.sdk.io.gcp.spanner.ReadOperation;
import org.apache.beam.sdk.io.gcp.spanner.SpannerConfig;
import org.apache.beam.sdk.io.gcp.spanner.Transaction;
import org.apache.beam.sdk.options.PipelineOptions;
import org.apache.beam.sdk.options.PipelineOptionsFactory;
import org.apache.beam.sdk.options.ValueProvider;
import org.apache.beam.sdk.transforms.Create;
import org.apache.beam.sdk.transforms.MapElements;
import org.apache.beam.sdk.transforms.PTransform;
import org.apache.beam.sdk.transforms.ParDo;
import org.apache.beam.sdk.transforms.SerializableFunction;
import org.apache.beam.sdk.transforms.View;
import org.apache.beam.sdk.values.PBegin;
import org.apache.beam.sdk.values.PCollection;
import org.apache.beam.sdk.values.PCollectionView;
import org.apache.beam.sdk.values.TypeDescriptors;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
/**
* Dataflow template which copies a Spanner table to a Text sink. It exports a Spanner table using
* <a href="https://cloud.google.com/spanner/docs/reads#read_data_in_parallel">Batch API</a>, which
* creates multiple workers in parallel for better performance. The result is written to a CSV file
* in Google Cloud Storage. The table schema file is saved in json format along with the exported
* table.
*
* <p>Schema file sample: { "id":"INT64", "name":"STRING(MAX)" }
*
* <p>A sample run:
*
* <pre>
* mvn compile exec:java \
* -Dexec.mainClass=com.google.cloud.teleport.templates.SpannerToText \
* -Dexec.args="--runner=DataflowRunner \
* --spannerProjectId=projectId \
* --gcpTempLocation=gs://gsTmpLocation \
* --spannerInstanceId=instanceId \
* --spannerDatabaseId=databaseId \
* --spannerTable=table_name \
* --spannerSnapshotTime=snapshot_time \
* --textWritePrefix=gcsOutputPath"
* </pre>
*/
@Template(
name = "Spanner_to_GCS_Text",
category = TemplateCategory.BATCH,
displayName = "Cloud Spanner to Text Files on Cloud Storage",
description =
"A pipeline which reads in Cloud Spanner table and writes it to Cloud Storage as CSV text files.",
optionsClass = SpannerToTextOptions.class,
contactInformation = "https://cloud.google.com/support")
public class SpannerToText {
private static final Logger LOG = LoggerFactory.getLogger(SpannerToText.class);
/** Custom PipelineOptions. */
public interface SpannerToTextOptions
extends PipelineOptions, SpannerReadOptions, FilesystemWriteOptions {
@TemplateParameter.GcsWriteFolder(
order = 1,
optional = true,
description = "Cloud Storage temp directory for storing CSV files",
helpText = "The Cloud Storage path where the temporary CSV files can be stored.",
example = "gs://your-bucket/your-path")
ValueProvider<String> getCsvTempDirectory();
@SuppressWarnings("unused")
void setCsvTempDirectory(ValueProvider<String> value);
@TemplateParameter.Enum(
order = 2,
enumOptions = {"LOW", "MEDIUM", "HIGH"},
optional = true,
description = "Priority for Spanner RPC invocations",
helpText =
"The request priority for Cloud Spanner calls. The value must be one of: [HIGH,MEDIUM,LOW].")
ValueProvider<RpcPriority> getSpannerPriority();
void setSpannerPriority(ValueProvider<RpcPriority> value);
}
/**
* Runs a pipeline which reads in Records from Spanner, and writes the CSV to TextIO sink.
*
* @param args arguments to the pipeline
*/
public static void main(String[] args) {
LOG.info("Starting pipeline setup");
PipelineOptionsFactory.register(SpannerToTextOptions.class);
SpannerToTextOptions options =
PipelineOptionsFactory.fromArgs(args).withValidation().as(SpannerToTextOptions.class);
FileSystems.setDefaultPipelineOptions(options);
Pipeline pipeline = Pipeline.create(options);
SpannerConfig spannerConfig =
SpannerConfig.create()
.withHost(options.getSpannerHost())
.withProjectId(options.getSpannerProjectId())
.withInstanceId(options.getSpannerInstanceId())
.withDatabaseId(options.getSpannerDatabaseId())
.withRpcPriority(options.getSpannerPriority());
PTransform<PBegin, PCollection<ReadOperation>> spannerExport =
SpannerConverters.ExportTransformFactory.create(
options.getSpannerTable(),
spannerConfig,
options.getTextWritePrefix(),
options.getSpannerSnapshotTime());
/* CreateTransaction and CreateTransactionFn classes in LocalSpannerIO
* only take a timestamp object for exact staleness which works when
* parameters are provided during template compile time. They do not work with
* a Timestamp valueProvider which can take parameters at runtime. Hence a new
* ParDo class CreateTransactionFnWithTimestamp had to be created for this
* purpose.
*/
PCollectionView<Transaction> tx =
pipeline
.apply("Setup for Transaction", Create.of(1))
.apply(
"Create transaction",
ParDo.of(
new CreateTransactionFnWithTimestamp(
spannerConfig, options.getSpannerSnapshotTime())))
.apply("As PCollectionView", View.asSingleton());
PCollection<String> csv =
pipeline
.apply("Create export", spannerExport)
// We need to use LocalSpannerIO.readAll() instead of LocalSpannerIO.read()
// because ValueProvider parameters such as table name required for
// LocalSpannerIO.read() can be read only inside DoFn but LocalSpannerIO.read() is of
// type PTransform<PBegin, Struct>, which prevents prepending it with DoFn that reads
// these parameters at the pipeline execution time.
.apply(
"Read all records",
LocalSpannerIO.readAll().withTransaction(tx).withSpannerConfig(spannerConfig))
.apply(
"Struct To Csv",
MapElements.into(TypeDescriptors.strings())
.via(struct -> (new SpannerConverters.StructCsvPrinter()).print(struct)));
ValueProvider<ResourceId> tempDirectoryResource =
ValueProvider.NestedValueProvider.of(
eitherOrValueProvider(options.getCsvTempDirectory(), options.getTextWritePrefix()),
(SerializableFunction<String, ResourceId>) s -> FileSystems.matchNewResource(s, true));
csv.apply(
"Write to storage",
TextIO.write()
.to(options.getTextWritePrefix())
.withSuffix(".csv")
.withTempDirectory(tempDirectoryResource));
pipeline.run();
LOG.info("Completed pipeline setup");
}
}
