/// <exception cref="System.IO.IOException"/>
public override void Commit()
{
fs.Rename(tmpOutputPath, outputPath);
MergeManagerImpl.CompressAwarePath compressAwarePath = new MergeManagerImpl.CompressAwarePath
(outputPath, GetSize(), this.compressedSize);
merger.CloseOnDiskFile(compressAwarePath);
}
public virtual void CloseOnDiskFile(MergeManagerImpl.CompressAwarePath file)
{
lock (this)
{
onDiskMapOutputs.AddItem(file);
if (onDiskMapOutputs.Count >= (2 * ioSortFactor - 1))
{
onDiskMerger.StartMerge(onDiskMapOutputs);
}
}
}
/// <exception cref="System.IO.IOException"/>
/// <exception cref="Sharpen.URISyntaxException"/>
/// <exception cref="System.Exception"/>
public virtual void TestOnDiskMerger()
{
JobConf jobConf = new JobConf();
int SortFactor = 5;
jobConf.SetInt(MRJobConfig.IoSortFactor, SortFactor);
MapOutputFile mapOutputFile = new MROutputFiles();
FileSystem fs = FileSystem.GetLocal(jobConf);
MergeManagerImpl <IntWritable, IntWritable> manager = new MergeManagerImpl <IntWritable
, IntWritable>(null, jobConf, fs, null, null, null, null, null, null, null, null
, null, null, mapOutputFile);
MergeThread <MapOutput <IntWritable, IntWritable>, IntWritable, IntWritable> onDiskMerger
= (MergeThread <MapOutput <IntWritable, IntWritable>, IntWritable, IntWritable>)Whitebox
.GetInternalState(manager, "onDiskMerger");
int mergeFactor = (int)Whitebox.GetInternalState(onDiskMerger, "mergeFactor");
// make sure the io.sort.factor is set properly
NUnit.Framework.Assert.AreEqual(mergeFactor, SortFactor);
// Stop the onDiskMerger thread so that we can intercept the list of files
// waiting to be merged.
onDiskMerger.Suspend();
//Send the list of fake files waiting to be merged
Random rand = new Random();
for (int i = 0; i < 2 * SortFactor; ++i)
{
Path path = new Path("somePath");
MergeManagerImpl.CompressAwarePath cap = new MergeManagerImpl.CompressAwarePath(path
, 1l, rand.Next());
manager.CloseOnDiskFile(cap);
}
//Check that the files pending to be merged are in sorted order.
List <IList <MergeManagerImpl.CompressAwarePath> > pendingToBeMerged = (List <IList <MergeManagerImpl.CompressAwarePath
> >)Whitebox.GetInternalState(onDiskMerger, "pendingToBeMerged");
NUnit.Framework.Assert.IsTrue("No inputs were added to list pending to merge", pendingToBeMerged
.Count > 0);
for (int i_1 = 0; i_1 < pendingToBeMerged.Count; ++i_1)
{
IList <MergeManagerImpl.CompressAwarePath> inputs = pendingToBeMerged[i_1];
for (int j = 1; j < inputs.Count; ++j)
{
NUnit.Framework.Assert.IsTrue("Not enough / too many inputs were going to be merged"
, inputs.Count > 0 && inputs.Count <= SortFactor);
NUnit.Framework.Assert.IsTrue("Inputs to be merged were not sorted according to size: "
, inputs[j].GetCompressedSize() >= inputs[j - 1].GetCompressedSize());
}
}
}
public override int CompareTo(object obj)
{
if (obj is MergeManagerImpl.CompressAwarePath)
{
MergeManagerImpl.CompressAwarePath compPath = (MergeManagerImpl.CompressAwarePath
)obj;
if (this.compressedSize < compPath.GetCompressedSize())
{
return(-1);
}
else
{
if (this.GetCompressedSize() > compPath.GetCompressedSize())
{
return(1);
}
}
}
// Not returning 0 here so that objects with the same size (but
// different paths) are still added to the TreeSet.
return(base.CompareTo(obj));
}
/// <exception cref="System.IO.IOException"/>
public override void Merge(IList <InMemoryMapOutput <K, V> > inputs)
{
if (inputs == null || inputs.Count == 0)
{
return;
}
//name this output file same as the name of the first file that is
//there in the current list of inmem files (this is guaranteed to
//be absent on the disk currently. So we don't overwrite a prev.
//created spill). Also we need to create the output file now since
//it is not guaranteed that this file will be present after merge
//is called (we delete empty files as soon as we see them
//in the merge method)
//figure out the mapId
TaskAttemptID mapId = inputs[0].GetMapId();
TaskID mapTaskId = mapId.GetTaskID();
IList <Merger.Segment <K, V> > inMemorySegments = new AList <Merger.Segment <K, V> >();
long mergeOutputSize = this._enclosing.CreateInMemorySegments(inputs, inMemorySegments
, 0);
int noInMemorySegments = inMemorySegments.Count;
Path outputPath = this._enclosing.mapOutputFile.GetInputFileForWrite(mapTaskId, mergeOutputSize
).Suffix(Org.Apache.Hadoop.Mapred.Task.MergedOutputPrefix);
FSDataOutputStream @out = CryptoUtils.WrapIfNecessary(this._enclosing.jobConf, this
._enclosing.rfs.Create(outputPath));
IFile.Writer <K, V> writer = new IFile.Writer <K, V>(this._enclosing.jobConf, @out,
(Type)this._enclosing.jobConf.GetMapOutputKeyClass(), (Type)this._enclosing.jobConf
.GetMapOutputValueClass(), this._enclosing.codec, null, true);
RawKeyValueIterator rIter = null;
MergeManagerImpl.CompressAwarePath compressAwarePath;
try
{
MergeManagerImpl.Log.Info("Initiating in-memory merge with " + noInMemorySegments
+ " segments...");
rIter = Merger.Merge(this._enclosing.jobConf, this._enclosing.rfs, (Type)this._enclosing
.jobConf.GetMapOutputKeyClass(), (Type)this._enclosing.jobConf.GetMapOutputValueClass
(), inMemorySegments, inMemorySegments.Count, new Path(this._enclosing.reduceId.
ToString()), (RawComparator <K>) this._enclosing.jobConf.GetOutputKeyComparator(),
this._enclosing.reporter, this._enclosing.spilledRecordsCounter, null, null);
if (null == this._enclosing.combinerClass)
{
Merger.WriteFile(rIter, writer, this._enclosing.reporter, this._enclosing.jobConf
);
}
else
{
this._enclosing.combineCollector.SetWriter(writer);
this._enclosing.CombineAndSpill(rIter, this._enclosing.reduceCombineInputCounter);
}
writer.Close();
compressAwarePath = new MergeManagerImpl.CompressAwarePath(outputPath, writer.GetRawLength
(), writer.GetCompressedLength());
MergeManagerImpl.Log.Info(this._enclosing.reduceId + " Merge of the " + noInMemorySegments
+ " files in-memory complete." + " Local file is " + outputPath + " of size " +
this._enclosing.localFS.GetFileStatus(outputPath).GetLen());
}
catch (IOException e)
{
//make sure that we delete the ondisk file that we created
//earlier when we invoked cloneFileAttributes
this._enclosing.localFS.Delete(outputPath, true);
throw;
}
// Note the output of the merge
this._enclosing.CloseOnDiskFile(compressAwarePath);
}
public virtual void TestInMemoryAndOnDiskMerger()
{
JobID jobId = new JobID("a", 0);
TaskAttemptID reduceId1 = new TaskAttemptID(new TaskID(jobId, TaskType.Reduce, 0)
, 0);
TaskAttemptID mapId1 = new TaskAttemptID(new TaskID(jobId, TaskType.Map, 1), 0);
TaskAttemptID mapId2 = new TaskAttemptID(new TaskID(jobId, TaskType.Map, 2), 0);
LocalDirAllocator lda = new LocalDirAllocator(MRConfig.LocalDir);
MergeManagerImpl <Text, Text> mergeManager = new MergeManagerImpl <Text, Text>(reduceId1
, jobConf, fs, lda, Reporter.Null, null, null, null, null, null, null, null, new
Progress(), new MROutputFiles());
// write map outputs
IDictionary <string, string> map1 = new SortedDictionary <string, string>();
map1["apple"] = "disgusting";
map1["carrot"] = "delicious";
IDictionary <string, string> map2 = new SortedDictionary <string, string>();
map1["banana"] = "pretty good";
byte[] mapOutputBytes1 = WriteMapOutput(conf, map1);
byte[] mapOutputBytes2 = WriteMapOutput(conf, map2);
InMemoryMapOutput <Text, Text> mapOutput1 = new InMemoryMapOutput <Text, Text>(conf
, mapId1, mergeManager, mapOutputBytes1.Length, null, true);
InMemoryMapOutput <Text, Text> mapOutput2 = new InMemoryMapOutput <Text, Text>(conf
, mapId2, mergeManager, mapOutputBytes2.Length, null, true);
System.Array.Copy(mapOutputBytes1, 0, mapOutput1.GetMemory(), 0, mapOutputBytes1.
Length);
System.Array.Copy(mapOutputBytes2, 0, mapOutput2.GetMemory(), 0, mapOutputBytes2.
Length);
// create merger and run merge
MergeThread <InMemoryMapOutput <Text, Text>, Text, Text> inMemoryMerger = mergeManager
.CreateInMemoryMerger();
IList <InMemoryMapOutput <Text, Text> > mapOutputs1 = new AList <InMemoryMapOutput <Text
, Text> >();
mapOutputs1.AddItem(mapOutput1);
mapOutputs1.AddItem(mapOutput2);
inMemoryMerger.Merge(mapOutputs1);
NUnit.Framework.Assert.AreEqual(1, mergeManager.onDiskMapOutputs.Count);
TaskAttemptID reduceId2 = new TaskAttemptID(new TaskID(jobId, TaskType.Reduce, 3)
, 0);
TaskAttemptID mapId3 = new TaskAttemptID(new TaskID(jobId, TaskType.Map, 4), 0);
TaskAttemptID mapId4 = new TaskAttemptID(new TaskID(jobId, TaskType.Map, 5), 0);
// write map outputs
IDictionary <string, string> map3 = new SortedDictionary <string, string>();
map3["apple"] = "awesome";
map3["carrot"] = "amazing";
IDictionary <string, string> map4 = new SortedDictionary <string, string>();
map4["banana"] = "bla";
byte[] mapOutputBytes3 = WriteMapOutput(conf, map3);
byte[] mapOutputBytes4 = WriteMapOutput(conf, map4);
InMemoryMapOutput <Text, Text> mapOutput3 = new InMemoryMapOutput <Text, Text>(conf
, mapId3, mergeManager, mapOutputBytes3.Length, null, true);
InMemoryMapOutput <Text, Text> mapOutput4 = new InMemoryMapOutput <Text, Text>(conf
, mapId4, mergeManager, mapOutputBytes4.Length, null, true);
System.Array.Copy(mapOutputBytes3, 0, mapOutput3.GetMemory(), 0, mapOutputBytes3.
Length);
System.Array.Copy(mapOutputBytes4, 0, mapOutput4.GetMemory(), 0, mapOutputBytes4.
Length);
// // create merger and run merge
MergeThread <InMemoryMapOutput <Text, Text>, Text, Text> inMemoryMerger2 = mergeManager
.CreateInMemoryMerger();
IList <InMemoryMapOutput <Text, Text> > mapOutputs2 = new AList <InMemoryMapOutput <Text
, Text> >();
mapOutputs2.AddItem(mapOutput3);
mapOutputs2.AddItem(mapOutput4);
inMemoryMerger2.Merge(mapOutputs2);
NUnit.Framework.Assert.AreEqual(2, mergeManager.onDiskMapOutputs.Count);
IList <MergeManagerImpl.CompressAwarePath> paths = new AList <MergeManagerImpl.CompressAwarePath
>();
IEnumerator <MergeManagerImpl.CompressAwarePath> iterator = mergeManager.onDiskMapOutputs
.GetEnumerator();
IList <string> keys = new AList <string>();
IList <string> values = new AList <string>();
while (iterator.HasNext())
{
MergeManagerImpl.CompressAwarePath next = iterator.Next();
ReadOnDiskMapOutput(conf, fs, next, keys, values);
paths.AddItem(next);
}
NUnit.Framework.Assert.AreEqual(keys, Arrays.AsList("apple", "banana", "carrot",
"apple", "banana", "carrot"));
NUnit.Framework.Assert.AreEqual(values, Arrays.AsList("awesome", "bla", "amazing"
, "disgusting", "pretty good", "delicious"));
mergeManager.Close();
mergeManager = new MergeManagerImpl <Text, Text>(reduceId2, jobConf, fs, lda, Reporter
.Null, null, null, null, null, null, null, null, new Progress(), new MROutputFiles
());
MergeThread <MergeManagerImpl.CompressAwarePath, Text, Text> onDiskMerger = mergeManager
.CreateOnDiskMerger();
onDiskMerger.Merge(paths);
NUnit.Framework.Assert.AreEqual(1, mergeManager.onDiskMapOutputs.Count);
keys = new AList <string>();
values = new AList <string>();
ReadOnDiskMapOutput(conf, fs, mergeManager.onDiskMapOutputs.GetEnumerator().Next(
), keys, values);
NUnit.Framework.Assert.AreEqual(keys, Arrays.AsList("apple", "apple", "banana", "banana"
, "carrot", "carrot"));
NUnit.Framework.Assert.AreEqual(values, Arrays.AsList("awesome", "disgusting", "pretty good"
, "bla", "amazing", "delicious"));
mergeManager.Close();
NUnit.Framework.Assert.AreEqual(0, mergeManager.inMemoryMapOutputs.Count);
NUnit.Framework.Assert.AreEqual(0, mergeManager.inMemoryMergedMapOutputs.Count);
NUnit.Framework.Assert.AreEqual(0, mergeManager.onDiskMapOutputs.Count);
}
