我想加入两个 PCollection（分别来自不同的输入）并按照此处描述的步骤实现，“加入 CoGroupByKey”部分： https ://cloud.google.com/dataflow/model/group-by-key
就我而言，我想加入 GeoIP 的“区块”信息和“位置”信息。因此，我将 Block 和 Location 定义为自定义类，然后编写如下：
final TupleTag<Block> t1 = new TupleTag<Block>();
final TupleTag<Location> t2 = new TupleTag<Location>();
PCollection<KV<Long, CoGbkResult>> coGbkResultColl = KeyedPCollectionTuple.of(t1, kvGeoNameIDBlock)
.and(t2, kvGeoNameIDLocation).apply(CoGroupByKey.<Long>create());
键具有 Long 类型的值。我以为它已经完成了，但是当我运行时mvn compile，它会输出以下错误：
[ERROR] Failed to execute goal org.codehaus.mojo:exec-maven-plugin:1.4.0:java (default-cli) on project xxxx: An exception occured while executing the Java class. null: InvocationTargetException: Unable to return a default Coder for Extract GeoNameID-Block KV/ParMultiDo(ExtractGeoNameIDBlock).out0 [PCollection]. Correct one of the following root causes:
[ERROR] No Coder has been manually specified; you may do so using .setCoder().
[ERROR] Inferring a Coder from the CoderRegistry failed: Cannot provide coder for parameterized type org.apache.beam.sdk.values.KV<java.lang.Long, com.xxx.platform.geoip2.Block>: Unable to provide a Coder for com.xxx.platform.geoip2.Block.
[ERROR] Building a Coder using a registered CoderProvider failed.
[ERROR] See suppressed exceptions for detailed failures.
[ERROR] Using the default output Coder from the producing PTransform failed: Cannot provide coder for parameterized type org.apache.beam.sdk.values.KV<java.lang.Long, com.xxx.platform.geoip2.Block>: Unable to provide a Coder for com.xxx.platform.geoip2.Block.
输出错误的确切 DoFn 是ExtractGeoNameIDBlock，它只是创建其键（要连接）和自身的键值对。
// ExtractGeoNameIDBlock creates KV collection while reading from block CSV
static class ExtractGeoNameIDBlock extends DoFn<String, KV<Long, Block>> {
private static final long serialVersionUID = 1L;
@ProcessElement
public void processElement(ProcessContext c) throws Exception {
String line = c.element();
if (!line.startsWith("network,")) { // exclude headerline
Block b = new Block();
b.loadFromCsvLine(line);
if (b.getGeonameId() != null) {
c.output(KV.of(b.getGeonameId(), b));
}
}
}
}
loadFromCsvLine只需解析 CSV 行，将字段转换为每个相应的类型并分配给它的私有字段。
所以看起来我需要为我的自定义类设置一些编码器才能使其工作。我找到了一个引用编码器的文档，但仍然不确定如何实现我的。 https://cloud.google.com/dataflow/model/data-encoding
有没有我可以遵循的真实示例来为我的自定义类创建自定义编码器？
[更新 13:02 09/26/2017] 我添加了
CoderRegistry cr = p.getCoderRegistry();
cr.registerCoderForClass(Block.class, AvroCoder.of(Block.class));
然后得到一个错误
java.lang.NullPointerException: in com.xxx.platform.geoip2.Block in long null of long in field representedCountryGeonameId of com.xxx.platform.geoip2.Block
[更新 14:05 09/26/2017] 我改变了这样的实现：
@DefaultCoder(AvroCoder.class)
public class Block {
private static final Logger LOG = LoggerFactory.getLogger(Block.class);
@Nullable
public String network;
@Nullable
public Long registeredCountryGeonameId;
:
:
（将@Nullable 设置为所有属性）
但仍然出现此错误：
(22eeaf3dfb26f8cc): java.lang.RuntimeException: org.apache.beam.sdk.coders.CoderException: cannot encode a null Long
at com.google.cloud.dataflow.worker.SimpleParDoFn$1.output(SimpleParDoFn.java:191)
at org.apache.beam.runners.core.SimpleDoFnRunner.outputWindowedValue(SimpleDoFnRunner.java:211)
at org.apache.beam.runners.core.SimpleDoFnRunner.access$700(SimpleDoFnRunner.java:66)
at org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:436)
at org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:424)
at org.apache.beam.sdk.transforms.join.CoGroupByKey$ConstructUnionTableFn.processElement(CoGroupByKey.java:185)
Caused by: org.apache.beam.sdk.coders.CoderException: cannot encode a null Long
at org.apache.beam.sdk.coders.VarLongCoder.encode(VarLongCoder.java:51)
at org.apache.beam.sdk.coders.VarLongCoder.encode(VarLongCoder.java:35)
at org.apache.beam.sdk.coders.Coder.encode(Coder.java:135)
at com.google.cloud.dataflow.worker.ShuffleSink$ShuffleSinkWriter.encodeToChunk(ShuffleSink.java:320)
at com.google.cloud.dataflow.worker.ShuffleSink$ShuffleSinkWriter.add(ShuffleSink.java:216)
at com.google.cloud.dataflow.worker.ShuffleSink$ShuffleSinkWriter.add(ShuffleSink.java:178)
at com.google.cloud.dataflow.worker.util.common.worker.WriteOperation.process(WriteOperation.java:80)
at com.google.cloud.dataflow.worker.util.common.worker.OutputReceiver.process(OutputReceiver.java:52)
at com.google.cloud.dataflow.worker.ReifyTimestampAndWindowsParDoFnFactory$ReifyTimestampAndWindowsParDoFn.processElement(ReifyTimestampAndWindowsParDoFnFactory.java:68)
at com.google.cloud.dataflow.worker.util.common.worker.ParDoOperation.process(ParDoOperation.java:48)
at com.google.cloud.dataflow.worker.util.common.worker.OutputReceiver.process(OutputReceiver.java:52)
at com.google.cloud.dataflow.worker.SimpleParDoFn$1.output(SimpleParDoFn.java:183)
at org.apache.beam.runners.core.SimpleDoFnRunner.outputWindowedValue(SimpleDoFnRunner.java:211)
at org.apache.beam.runners.core.SimpleDoFnRunner.access$700(SimpleDoFnRunner.java:66)
at org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:436)
at org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:424)
at org.apache.beam.sdk.transforms.join.CoGroupByKey$ConstructUnionTableFn.processElement(CoGroupByKey.java:185)
at org.apache.beam.sdk.transforms.join.CoGroupByKey$ConstructUnionTableFn$DoFnInvoker.invokeProcessElement(Unknown Source)
at org.apache.beam.runners.core.SimpleDoFnRunner.invokeProcessElement(SimpleDoFnRunner.java:177)
at org.apache.beam.runners.core.SimpleDoFnRunner.processElement(SimpleDoFnRunner.java:141)
at com.google.cloud.dataflow.worker.SimpleParDoFn.processElement(SimpleParDoFn.java:233)
at com.google.cloud.dataflow.worker.util.common.worker.ParDoOperation.process(ParDoOperation.java:48)
at com.google.cloud.dataflow.worker.util.common.worker.OutputReceiver.process(OutputReceiver.java:52)
at com.google.cloud.dataflow.worker.SimpleParDoFn$1.output(SimpleParDoFn.java:183)
at org.apache.beam.runners.core.SimpleDoFnRunner.outputWindowedValue(SimpleDoFnRunner.java:211)
at org.apache.beam.runners.core.SimpleDoFnRunner.access$700(SimpleDoFnRunner.java:66)
at org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:436)
at org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:424)
at com.bandainamcoent.platform.GeoIpPopulateTable$ExtractGeoNameIDBlock.processElement(GeoIpPopulateTable.java:79)
at com.bandainamcoent.platform.GeoIpPopulateTable$ExtractGeoNameIDBlock$DoFnInvoker.invokeProcessElement(Unknown Source)
at org.apache.beam.runners.core.SimpleDoFnRunner.invokeProcessElement(SimpleDoFnRunner.java:177)
at org.apache.beam.runners.core.SimpleDoFnRunner.processElement(SimpleDoFnRunner.java:141)
at com.google.cloud.dataflow.worker.SimpleParDoFn.processElement(SimpleParDoFn.java:233)
at com.google.cloud.dataflow.worker.util.common.worker.ParDoOperation.process(ParDoOperation.java:48)
at com.google.cloud.dataflow.worker.util.common.worker.OutputReceiver.process(OutputReceiver.java:52)
at com.google.cloud.dataflow.worker.util.common.worker.ReadOperation.runReadLoop(ReadOperation.java:187)
at com.google.cloud.dataflow.worker.util.common.worker.ReadOperation.start(ReadOperation.java:148)
at com.google.cloud.dataflow.worker.util.common.worker.MapTaskExecutor.execute(MapTaskExecutor.java:68)
at com.google.cloud.dataflow.worker.DataflowWorker.executeWork(DataflowWorker.java:336)
at com.google.cloud.dataflow.worker.DataflowWorker.doWork(DataflowWorker.java:294)
at com.google.cloud.dataflow.worker.DataflowWorker.getAndPerformWork(DataflowWorker.java:244)
at com.google.cloud.dataflow.worker.DataflowBatchWorkerHarness$WorkerThread.doWork(DataflowBatchWorkerHarness.java:135)
at com.google.cloud.dataflow.worker.DataflowBatchWorkerHarness$WorkerThread.call(DataflowBatchWorkerHarness.java:115)
at com.google.cloud.dataflow.worker.DataflowBatchWorkerHarness$WorkerThread.call(DataflowBatchWorkerHarness.java:102)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
谢谢。