我有date
&分区的镶木地板数据。 hour
,文件夹结构:
events_v3
-- event_date=2015-01-01
-- event_hour=2015-01-1
-- part10000.parquet.gz
-- event_date=2015-01-02
-- event_hour=5
-- part10000.parquet.gz
我已经通过spark创建了一个表raw_events
,但是当我尝试查询时,它会扫描所有目录的页脚,这会减慢初始查询的速度,即使我只查询一天的数据。 / p>
查询:
select * from raw_events where event_date='2016-01-01'
日志:
App > 16/09/15 03:14:03 main INFO HadoopFsRelation: Listing leaf files and directories in parallel under: s3a://bucket/events_v3/
然后它产生350个任务,因为有350天的数据。
我已禁用schemaMerge
,并且还指定了要读取的架构,因此它可以转到我正在查看的分区,为什么要打印所有叶子文件?
列出具有2个执行程序的叶子文件需要10分钟,查询实际执行需要20秒
代码示例:
val sparkSession = org.apache.spark.sql.SparkSession.builder.getOrCreate()
val df = sparkSession.read.option("mergeSchema","false").format("parquet").load("s3a://bucket/events_v3")
df.createOrReplaceTempView("temp_events")
sparkSession.sql(
"""
|select verb,count(*) from temp_events where event_date = "2016-01-01" group by verb
""".stripMargin).show()
答案 0 :(得分:8)
一旦给出了一个要从中读取的目录,就会调用listLeafFiles
(org / apache / spark / sql / execution / datasources / fileSourceInterfaces.scala)。这反过来调用fs.listStatus
进行api调用以获取文件和目录列表。现在,对于每个目录,再次调用此方法。这是递归的,直到没有目录。这种设计在HDFS系统中运行良好。但是在s3中工作不好,因为列表文件是RPC调用。其他S3支持通过前缀获取所有文件,这正是我们所需要的。
因此,例如,如果我们的目录结构具有1年的数据,每个目录为小时和10个子目录,我们将有,365 * 24 * 10 = 87k api调用,这可以减少到138 api调用那里只有137000个文件。每个s3 api调用返回1000个文件。
代码:
org/apache/hadoop/fs/s3a/S3AFileSystem.java
public FileStatus[] listStatusRecursively(Path f) throws FileNotFoundException,
IOException {
String key = pathToKey(f);
if (LOG.isDebugEnabled()) {
LOG.debug("List status for path: " + f);
}
final List<FileStatus> result = new ArrayList<FileStatus>();
final FileStatus fileStatus = getFileStatus(f);
if (fileStatus.isDirectory()) {
if (!key.isEmpty()) {
key = key + "/";
}
ListObjectsRequest request = new ListObjectsRequest();
request.setBucketName(bucket);
request.setPrefix(key);
request.setMaxKeys(maxKeys);
if (LOG.isDebugEnabled()) {
LOG.debug("listStatus: doing listObjects for directory " + key);
}
ObjectListing objects = s3.listObjects(request);
statistics.incrementReadOps(1);
while (true) {
for (S3ObjectSummary summary : objects.getObjectSummaries()) {
Path keyPath = keyToPath(summary.getKey()).makeQualified(uri, workingDir);
// Skip over keys that are ourselves and old S3N _$folder$ files
if (keyPath.equals(f) || summary.getKey().endsWith(S3N_FOLDER_SUFFIX)) {
if (LOG.isDebugEnabled()) {
LOG.debug("Ignoring: " + keyPath);
}
continue;
}
if (objectRepresentsDirectory(summary.getKey(), summary.getSize())) {
result.add(new S3AFileStatus(true, true, keyPath));
if (LOG.isDebugEnabled()) {
LOG.debug("Adding: fd: " + keyPath);
}
} else {
result.add(new S3AFileStatus(summary.getSize(),
dateToLong(summary.getLastModified()), keyPath,
getDefaultBlockSize(f.makeQualified(uri, workingDir))));
if (LOG.isDebugEnabled()) {
LOG.debug("Adding: fi: " + keyPath);
}
}
}
for (String prefix : objects.getCommonPrefixes()) {
Path keyPath = keyToPath(prefix).makeQualified(uri, workingDir);
if (keyPath.equals(f)) {
continue;
}
result.add(new S3AFileStatus(true, false, keyPath));
if (LOG.isDebugEnabled()) {
LOG.debug("Adding: rd: " + keyPath);
}
}
if (objects.isTruncated()) {
if (LOG.isDebugEnabled()) {
LOG.debug("listStatus: list truncated - getting next batch");
}
objects = s3.listNextBatchOfObjects(objects);
statistics.incrementReadOps(1);
} else {
break;
}
}
} else {
if (LOG.isDebugEnabled()) {
LOG.debug("Adding: rd (not a dir): " + f);
}
result.add(fileStatus);
}
return result.toArray(new FileStatus[result.size()]);
}
/org/apache/spark/sql/execution/datasources/fileSourceInterfaces.scala
def listLeafFiles(fs: FileSystem, status: FileStatus, filter: PathFilter): Array[FileStatus] = {
logTrace(s"Listing ${status.getPath}")
val name = status.getPath.getName.toLowerCase
if (shouldFilterOut(name)) {
Array.empty[FileStatus]
}
else {
val statuses = {
val stats = if(fs.isInstanceOf[S3AFileSystem]){
logWarning("Using Monkey patched version of list status")
println("Using Monkey patched version of list status")
val a = fs.asInstanceOf[S3AFileSystem].listStatusRecursively(status.getPath)
a
// Array.empty[FileStatus]
}
else{
val (dirs, files) = fs.listStatus(status.getPath).partition(_.isDirectory)
files ++ dirs.flatMap(dir => listLeafFiles(fs, dir, filter))
}
if (filter != null) stats.filter(f => filter.accept(f.getPath)) else stats
}
// statuses do not have any dirs.
statuses.filterNot(status => shouldFilterOut(status.getPath.getName)).map {
case f: LocatedFileStatus => f
// NOTE:
//
// - Although S3/S3A/S3N file system can be quite slow for remote file metadata
// operations, calling `getFileBlockLocations` does no harm here since these file system
// implementations don't actually issue RPC for this method.
//
// - Here we are calling `getFileBlockLocations` in a sequential manner, but it should not
// be a big deal since we always use to `listLeafFilesInParallel` when the number of
// paths exceeds threshold.
case f => createLocatedFileStatus(f, fs.getFileBlockLocations(f, 0, f.getLen))
}
}
}
答案 1 :(得分:2)
为了澄清Gaurav的答案,该代码剪辑来自Hadoop branch-2,可能直到Hadoop 2.9才会浮出水面(见HADOOP-13208);并且有人需要更新Spark以使用该功能(使用HDFS不会损害代码,只是赢得了显示任何加速)。
需要考虑的一件事是:什么能成为对象商店的良好文件布局。
最后,Apache Hadoop,Apache Spark和相关项目都是开源的。欢迎捐款。这不仅仅是代码,它的文档,测试,以及针对这些性能的东西,还可以针对您的实际数据集进行测试。甚至向我们提供有关导致问题的原因(以及您的数据集布局)的详细信息也很有趣。