我正在尝试向HDInsight集群提交MapReduce作业。在我的工作中,我没有写减少部分,因为我不想减少任何东西。我想做的只是解析每个文件名并将值附加到文件中的每一行。这样我就可以在文件中拥有所需的所有数据。
我的代码是
using Microsoft.Hadoop.MapReduce;
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace GetMetaDataFromFileName
{
class Program
{
static void Main(string[] args)
{
var hadoop = connectAzure();
//Temp Workaround to Env Variables
Environment.SetEnvironmentVariable("HADOOP_HOME", @"c:hadoop");
Environment.SetEnvironmentVariable("Java_HOME", @"c:hadoopjvm");
var result = hadoop.MapReduceJob.ExecuteJob<MetaDataGetterJob>();
}
static IHadoop connectAzure()
{
//TODO: Update credentials and other information
return Hadoop.Connect(
new Uri("https://sampleclustername.azurehdinsight.net//"),
"admin",
"Hadoop",
"password",
"blobstoragename.blob.core.windows.net", //Storage Account that Log files exists
"AccessKeySample", //Storage Account Access Key
"logs", //Container Name
true
);
}
//Hadoop Mapper
public class MetaDataGetter : MapperBase
{
public override void Map(string inputLine, MapperContext context)
{
try
{
//Get the meta data from name of the file
string[] _fileMetaData = context.InputFilename.Split('_');
string _PublicIP = _fileMetaData[0].Trim();
string _PhysicalAdapterMAC = _fileMetaData[1].Trim();
string _BootID = _fileMetaData[2].Trim();
string _ServerUploadTime = _fileMetaData[3].Trim();
string _LogType = _fileMetaData[4].Trim();
string _MachineUpTime = _fileMetaData[5].Trim();
//Generate CSV portion
string _RowHeader = string.Format("{0},{1},{2},{3},{4},{5},", _PublicIP, _PhysicalAdapterMAC, _BootID, _ServerUploadTime, _LogType, _MachineUpTime);
//TODO: Append _RowHeader to every row in the file.
context.EmitLine(_RowHeader + inputLine);
}
catch(ArgumentException ex)
{
return;
}
}
}
//Hadoop Job Definition
public class MetaDataGetterJob : HadoopJob<MetaDataGetter>
{
public override HadoopJobConfiguration Configure(ExecutorContext context)
{
//Initiate the job config
HadoopJobConfiguration config = new HadoopJobConfiguration();
config.InputPath = "asv://logs@sample.blob.core.windows.net/Input";
config.OutputFolder = "asv://logs@sample.blob.core.windows.net/Output";
config.DeleteOutputFolder = true;
return config;
}
}
}
}
通常你认为500(服务器错误)的原因是什么?我是否提供了错误的凭据?实际上我并不真正了解Hadoop.Connect方法中用户名和HadoopUser参数之间的区别?
谢谢
我过去遇到过大致相同的问题(无法使用 BadGateway 响应向集群提交 hive 作业)。我已经联系了支持团队,在我的情况下,问题出在头节点的内存泄漏中,这意味着问题不在客户端,似乎是遗传的Hadoop问题。
我已经通过重新部署集群解决了这些问题。您是否尝试过提交其他工作(简单的工作)?如果是这样,那么我建议与 azure 支持团队联系,或者只是重新部署群集,如果这对你来说不痛苦的话。