Flink可查询状态错误

Question

我试图在Flink（版本1.4.2）上使用可查询状态，但不幸的是，我一直收到以下错误：

INFO  my.test.flink.QueryableState  - Params are a96438fa12879b7598c9cf32684e2669, kafka-cluster_jobmanager_1, 6123
INFO  my.test.flink.QueryableState  - Before the call java.util.concurrent.CompletableFuture@26aa12dd[Not completed]
java.util.concurrent.ExecutionException: java.lang.IndexOutOfBoundsException: readerIndex(0) + length(4) exceeds writerIndex(0): PooledUnsafeDirectByteBuf(ridx: 0, widx: 0, cap: 0)
        at java.util.concurrent.CompletableFuture.reportGet(CompletableFuture.java:357)
        at java.util.concurrent.CompletableFuture.get(CompletableFuture.java:1895)
        at my.test.flink.QueryableState.main(QueryableState.java:67)
Caused by: java.lang.IndexOutOfBoundsException: readerIndex(0) + length(4) exceeds writerIndex(0): PooledUnsafeDirectByteBuf(ridx: 0, widx: 0, cap: 0)
        at org.apache.flink.shaded.netty4.io.netty.buffer.AbstractByteBuf.checkReadableBytes(AbstractByteBuf.java:1166)
        at org.apache.flink.shaded.netty4.io.netty.buffer.AbstractByteBuf.readInt(AbstractByteBuf.java:619)
        at org.apache.flink.queryablestate.network.messages.MessageSerializer.deserializeHeader(MessageSerializer.java:231)
        at org.apache.flink.queryablestate.network.ClientHandler.channelRead(ClientHandler.java:76)
        at org.apache.flink.shaded.netty4.io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:339)
        at org.apache.flink.shaded.netty4.io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:324)
        at org.apache.flink.shaded.netty4.io.netty.channel.ChannelInboundHandlerAdapter.channelRead(ChannelInboundHandlerAdapter.java:86)
        at org.apache.flink.shaded.netty4.io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:339)
        at org.apache.flink.shaded.netty4.io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:324)
        at org.apache.flink.shaded.netty4.io.netty.handler.codec.ByteToMessageDecoder.channelRead(ByteToMessageDecoder.java:242)
        at org.apache.flink.shaded.netty4.io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:339)
        at org.apache.flink.shaded.netty4.io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:324)
        at org.apache.flink.shaded.netty4.io.netty.channel.DefaultChannelPipeline.fireChannelRead(DefaultChannelPipeline.java:847)
        at org.apache.flink.shaded.netty4.io.netty.channel.nio.AbstractNioByteChannel$NioByteUnsafe.read(AbstractNioByteChannel.java:131)
        at org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:511)
        at org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:468)
        at org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:382)
        at org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:354)
        at org.apache.flink.shaded.netty4.io.netty.util.concurrent.SingleThreadEventExecutor$2.run(SingleThreadEventExecutor.java:111)
        at java.lang.Thread.run(Thread.java:745)

在客户端，我使用flink-queryable-state-client-java_2_11.jar可查询客户端的相关代码部分是

QueryableStateClient client = new QueryableStateClient(jobManagerHost, jobManagerPort);

TypeInformation<MyEvent> typeInformation = TypeInformation.of(new TypeHint<MyEvent>() {});
ListStateDescriptor<MyEvent> descriptor = new ListStateDescriptor<MyEvent>("myEvents",
                 typeInformation.createSerializer(new ExecutionConfig()));

CompletableFuture<ListState<MyEvent>> resultFuture =
                        client.getKvState(JobID.fromHexString(jobIdParam),"myEvents", "1", 
                        BasicTypeInfo.STRING_TYPE_INFO , descriptor );

logger.info("Before the call " + resultFuture);
try {
         logger.info("Finished"+ resultFuture.get());
 } catch(Exception ex) {
         ex.printStackTrace();
 }

最后，在Flink上运行的作业配置了一个ListState，如下所示。请注意，数据在ListState上由String键控

        TypeInformation<MyEvent> typeInformation = TypeInformation.of(new TypeHint<MyEvent>() {});
        ListStateDescriptor<MyEvent> eventState = 
                new ListStateDescriptor<MyEvent>("myEvents",typeInformation);
        eventState.setQueryable("myEvents");
        eventListState = getRuntimeContext().getListState(eventState);

在我看来，这似乎是一个序列化错误，但我不知道我需要做什么来修复它。有人知道上面的代码可能有什么问题吗？我错过什么了吗？

Answer 1

我在更新Flink 1.4的可查询状态演示时遇到了完全相同的问题。如果我没记错的话，重要的部分是正确处理可完成的未来——你不能直接调用get（）。

请参阅工作示例的代码，其关键部分如下所示：

try {

    CompletableFuture<FoldingState<BumpEvent, Long>> resultFuture =
      client.getKvState(jobId, EventCountJob.ITEM_COUNTS, key, 
      BasicTypeInfo.STRING_TYPE_INFO, countingState);

    resultFuture.thenAccept(response -> {
      try {
        Long count = response.get();
        // now we could do something with the value
      } catch (Exception e) {
        e.printStackTrace();
      }
    });

    resultFuture.get(5, TimeUnit.SECONDS);

} catch (Exception e) {
  e.printStackTrace();
}