Overview of Spark project

1. Overview of Spark project Presented by Yin Zhu (yinz@ust.hk) Materials from http://spark-project.org/documentation/ Hadoop in Practice by A. Holmes My demo code: https://bitbucket.org/blackswift/spark-example 25 March 2013

2. Outline Review of MapReduce(15’) Going Through Spark (NSDI’12) Slides (30’) Demo (15’)

3. Review of MapReduce • Pagerank implemented in Scala and Hadoop • Why Pagerank? • >More complicated than the “hello world” WordCount • >Widely used in search engines, very large scale input (the whole web graph!) • >Iterative algorithm (typical style for most numerical algorithms for data mining)

5. New score: 0.15 + 0.5*0.85 = 0.575

10. A functional implementation • defpagerank(links: Array[(UrlId, Array[UrlId])], numIters: Int):Map[UrlId, Double] = { • valn = links.size • varranks = (Array.fromFunction(i => (i, 1.0)) (n)).toMap// init: each node has rank 1.0 • for (iter <- 1 to numIters) { // Take some interactions • valcontrib = • links • .flatMap(node => { • valout_url = node._1 • valin_urls = node._2 • valscore = ranks(out_url) / in_urls.size // the score each outer link recieves • in_urls.map(in_url => (in_url, score) ) • } • ) • .groupBy(url_score => url_score._1) // group the (url, score) pairs by url • .map(url_scoregroup => // sum the score for each unique url • (url_scoregroup._1, url_scoregroup._2.foldLeft (0.0) ((sum,url_score) => sum+url_score._2))) • ranks = ranks.map(url_score => (url_score._1, if (contrib.contains(url_score._1)) 0.85 * contrib(url_score._1) + 0.15else url_score._2)) • } • ranks • } Map Reduce

11. Hadoop implementation • https://github.com/alexholmes/hadoop-book/tree/master/src/main/java/com/manning/hip/ch7/pagerank

12. Hadoop/MapReduce implementation • Fault tolerance • the result after each iteration is saved to Disk • Speed/Disk IO • disk IO is proportional to the # of iterations • ideally the link graph should be loaded for only once

13. Resilient Distributed Datasets A Fault-Tolerant Abstraction forIn-Memory Cluster Computing Matei Zaharia, MosharafChowdhury, Tathagata Das, Ankur Dave, Justin Ma, Murphy McCauley,Michael Franklin, Scott Shenker, Ion Stoica UC Berkeley UC BERKELEY

14. Motivation • MapReduce greatly simplified “big data” analysis on large, unreliable clusters • But as soon as it got popular, users wanted more: • More complex, multi-stage applications(e.g. iterative machine learning & graph processing) • More interactive ad-hoc queries Response: specialized frameworks for some of these apps (e.g. Pregel for graph processing)

15. Motivation • Complex apps and interactive queries both need one thing that MapReduce lacks: • Efficient primitives for data sharing In MapReduce, the only way to share data across jobs is stable storage  slow!

16. Examples HDFSread HDFSwrite HDFSread HDFSwrite iter. 1 iter. 2 . . . Input result 1 query 1 HDFSread result 2 query 2 query 3 result 3 Input . . . Slow due to replication and disk I/O,but necessary for fault tolerance

17. Goal: In-Memory Data Sharing iter. 1 iter. 2 . . . Input query 1 one-timeprocessing query 2 query 3 Input . . . 10-100× faster than network/disk, but how to get FT?

18. Challenge • How to design a distributed memory abstraction that is both fault-tolerant and efficient?

19. Solution: Resilient Distributed Datasets (RDDs) • Restricted form of distributed shared memory • Immutable, partitioned collections of records • Can only be built through coarse-grained deterministic transformations (map, filter, join, …) • Efficient fault recovery using lineage • Log one operation to apply to many elements • Recompute lost partitions on failure • No cost if nothing fails

20. Three core concepts of RDD • Transformations • define a new RDD from an existing RDD • Cache and Partitioner • Put the dataset into memory/other persist media, and specify the locations of the sub datasets • Actions • carry out the actual computation

21. RDD and its lazy transformations • RDD[T]: A sequenceof objects of type T • Transformations are lazy: https://github.com/mesos/spark/blob/master/core/src/main/scala/spark/PairRDDFunctions.scala

22. /** * Return an RDD of grouped items. */ defgroupBy[K:ClassManifest](f:T=>K,p:Partitioner):RDD[(K, Seq[T])]={ valcleanF=sc.clean(f) this.map(t=>(cleanF(t),t)).groupByKey(p) } /** * Group the values for each key in the RDD into a single sequence. Allows controlling the * partitioning of the resulting key-value pair RDD by passing a Partitioner. */ defgroupByKey(partitioner:Partitioner):RDD[(K, Seq[V])]={ defcreateCombiner(v:V)=ArrayBuffer(v) defmergeValue(buf:ArrayBuffer[V],v:V)=buf+=v defmergeCombiners(b1:ArrayBuffer[V],b2:ArrayBuffer[V])=b1++=b2 valbufs=combineByKey[ArrayBuffer[V]]( createCombiner_,mergeValue_,mergeCombiners_,partitioner) bufs.asInstanceOf[RDD[(K, Seq[V])]] }

23. Actions: carry out the actual computation https://github.com/mesos/spark/blob/master/core/src/main/scala/spark/SparkContext.scala runJob

24. Example • lines = spark.textFile(“hdfs://...”) • errors = lines.filter(_.startsWith(“ERROR”)) • messages = errors.map(_.split(‘\t’)(2)) • messages.persist() or .cache() messages.filter(_.contains(“foo”)).count messages.filter(_.contains(“bar”)).count

25. Task Scheduler for actions Dryad-like DAGs Pipelines functionswithin a stage Locality & data reuseaware Partitioning-awareto avoid shuffles B: A: G: Stage 1 groupBy F: D: C: map E: join Stage 2 union Stage 3 = cached data partition

26. RDD Recovery iter. 1 iter. 2 . . . Input query 1 one-timeprocessing query 2 query 3 Input . . .

27. Fault Recovery • RDDs track the graph of transformations that built them (their lineage) to rebuild lost data E.g.: messages = textFile(...).filter(_.contains(“error”)) .map(_.split(‘\t’)(2)) HadoopRDD FilteredRDD MappedRDD HadoopRDD path = hdfs://… FilteredRDD func = _.contains(...) MappedRDD func = _.split(…)

28. Fault Recovery Results Failure happens

29. Generality of RDDs • Despite their restrictions, RDDs can express surprisingly many parallel algorithms • These naturally apply the same operation to many items • Unify many current programming models • Data flow models:MapReduce, Dryad, SQL, … • Specialized models for iterative apps:BSP (Pregel), iterative MapReduce (Haloop), bulk incremental, … • Support new apps that these models don’t

30. Tradeoff Space Network bandwidth Memory bandwidth Fine Best for transactional workloads K-V stores, databases, RAMCloud Granularity of Updates Best for batch workloads HDFS RDDs Coarse Low High Write Throughput

31. Outline Spark programming interface Implementation Demo How people are using Spark

32. Spark Programming Interface • DryadLINQ-like API in the Scala language • Usable interactively from Scala interpreter • Provides: • Resilient distributed datasets (RDDs) • Operations on RDDs: transformations (build new RDDs), actions (compute and output results) • Control of each RDD’s partitioning (layout across nodes) and persistence (storage in RAM, on disk, etc)

33. Example: Log Mining Load error messages from a log into memory, then interactively search for various patterns Msgs. 1 Base RDD Transformed RDD lines = spark.textFile(“hdfs://...”) errors = lines.filter(_.startsWith(“ERROR”)) messages = errors.map(_.split(‘\t’)(2)) messages.persist() Worker results tasks Block 1 Master Action messages.filter(_.contains(“foo”)).count Msgs. 2 messages.filter(_.contains(“bar”)).count Worker Msgs. 3 Block 2 Worker Result: full-text search of Wikipedia in <1 sec (vs 20 sec for on-disk data) Result: scaled to 1 TB data in 5-7 sec(vs 170 sec for on-disk data) Block 3

34. Example: PageRank • 1. Start each page with a rank of 1 • 2. On each iteration, update each page’s rank to • Σi∈neighborsranki/ |neighborsi| links = // RDD of (url, neighbors) pairs ranks = // RDD of (url, rank) pairs for (i <- 1 to ITERATIONS) { ranks = links.join(ranks).flatMap { (url, (links, rank)) => links.map(dest => (dest, rank/links.size)) }.reduceByKey(_ + _) }

35. Optimizing Placement • links & ranks repeatedly joined • Can co-partition them (e.g. hash both on URL) to avoid shuffles • Can also use app knowledge, e.g., hash on DNS name • links = links.partitionBy( new URLPartitioner()) Links (url, neighbors) Ranks0 (url, rank) join Contribs0 reduce Ranks1 join Contribs2 reduce Ranks2 . . .

36. PageRank Performance

37. Implementation Runs on Mesos [NSDI 11]to share clusters w/ Hadoop Can read from any Hadoop input source (HDFS, S3, …) • No changes to Scala language or compiler • Reflection + bytecode analysis to correctly ship code • www.spark-project.org Spark Hadoop MPI … Mesos Node Node Node Node

38. Behavior with Insufficient RAM

39. Scalability Logistic Regression K-Means

40. Breaking Down the Speedup

41. Spark Operations

42. Demo

43. Open Source Community • 15 contributors, 5+ companies using Spark,3+ applications projects at Berkeley • User applications: • Data mining 40x faster than Hadoop (Conviva) • Exploratory log analysis (Foursquare) • Traffic prediction via EM (Mobile Millennium) • Twitter spam classification (Monarch) • DNA sequence analysis (SNAP) • . . .

44. Related Work RAMCloud, Piccolo, GraphLab, parallel DBs • Fine-grained writes requiring replication for resilience • Pregel, iterative MapReduce • Specialized models; can’t run arbitrary / ad-hoc queries • DryadLINQ, FlumeJava • Language-integrated “distributed dataset” API, but cannot share datasets efficiently across queries • Nectar [OSDI 10] • Automatic expression caching, but over distributed FS • PacMan [NSDI 12] • Memory cache for HDFS, but writes still go to network/disk

45. Conclusion • RDDs offer a simple and efficient programming model for a broad range of applications • Leverage the coarse-grained nature of many parallel algorithms for low-overhead recovery • Try it out at www.spark-project.org