Programming and Debugging Large- Scale Data Processing Workflows

Transcription

1 Programming and Debugging Large- Scale Data Processing Workflows Christopher Olston Google Research (work done at Yahoo! Research, with many colleagues)

2 Context Elaborate processing of large data sets e.g.: web search pre- processing cross- dataset linkage web informakon extrackon inges)on storage & processing serving

3 Context storage & processing workflow manager e.g. Nova dataflow programming framework e.g. Pig distributed sorkng & hashing e.g. Map- Reduce Overview scalable file system e.g. GFS Detail: Inspector Gadget, RubySky Debugging aides: Before: example data generator During: instrumentakon framework ANer: provenance metadata manager

4 Pig: A High- Level Dataflow Language & RunKme for Hadoop Web browsing sessions with happy endings. Visits = load /data/visits as (user, url, time);! Visits = foreach Visits generate user, Canonicalize(url), time;!! Pages = load /data/pages as (url, pagerank);!! VP = join Visits by url, Pages by url;! UserVisits = group VP by user;! Sessions = foreach UserVisits generate flatten(findsessions(*));! HappyEndings = filter Sessions by BestIsLast(*);!! store HappyEndings into '/data/happy_endings';!

5 vs. map- reduce: less code! "The [Hofmann PLSA E/M] algorithm was implemented in pig in lines of pig-latin statements. Took a lot less compared to what it took in implementing the algorithm in Map-Reduce Java. Exactly that's the reason I wanted to try it out in Pig. It took 3-4 days for me to write it, starting from learning pig. " " -- Prasenjit Mukherjee, Mahout project" /20 the lines of code Minutes /16 the development Kme Hadoop Pig Hadoop Pig performs on par with raw Hadoop

6 vs. SQL: step- by- step style; lower- level control "I much prefer writing in Pig [Latin] versus SQL. The step-by-step method of" creating a program in Pig [Latin] is much cleaner and simpler to use than the single block method of SQL. It is easier to keep track of what your variables are, and where you are in the process of analyzing your data. " " -- Jasmine Novak, Engineer, Yahoo!" "PIG seems to give the necessary parallel programming construct (FOREACH, FLATTEN, COGROUP.. etc) and also give sufficient control back to the programmer (which purely declarative approach like [SQL on Map-Reduce] doesnʼt). " " -- Ricky Ho, Adobe Software"

7 Conceptually: A Graph of Data TransformaKons Find users who tend to visit good pages. Load Visits(user, url, Kme) Load Pages(url, pagerank) Transform to (user, Canonicalize(url), Kme) Join url = url Group by user Transform to (user, Average(pagerank) as avgpr) Filter avgpr > 0.5

8 Load Visits(user, url, Kme) Load Pages(url, pagerank) Illustrated! Transform to (user, Canonicalize(url), Kme) (Amy, cnn.com, 8am) (Amy, hip:// 9am) (Fred, 11am) (Amy, 8am) (Amy, 9am) (Fred, 11am) Join url = url Group by user (Amy, 8am, 0.9) (Amy, 9am, 0.4) (Fred, 11am, 0.4) ( 0.9) ( 0.4) (Amy, { (Amy, 8am, 0.9), (Amy, 9am, 0.4) }) (Fred, { (Fred, 11am, 0.4) }) Transform to (user, Average(pagerank) as avgpr) ILLUSTRATE lets me check the output of my lengthy (Amy, 0.65) batch jobs and their (Fred, 0.4) custom functions without having to do a lengthy run of a long pipeline. [This feature] enables me to be productive. Filter " " avgpr > Russell Jurney, LinkedIn" (Amy, 0.65)

9 (Naïve Algorithm) Load Visits(user, url, Kme) Transform to (user, Canonicalize(url), Kme) (Amy, cnn.com, 8am) (Amy, hip:// 9am) (Fred, 11am) (Amy, 8am) (Amy, 9am) (Fred, 11am) Join url = url Group by user Transform to (user, Average(pagerank) as avgpr) Load Pages(url, pagerank) ( 0.9) ( 0.4) Filter avgpr > 0.5

10 Original UI Prototype (2008)

11 InteracKve Graphics Programming UI by Bret Victor, Apple Designer (2012) Creators need an immediate connection to what they create If you make a change you need to see the effect of that immediately. " " -- Bret Victor, Apple"

12 Pig Today Open- source (the Apache Pig Project) Dev./support/training by Cloudera, Hortonworks Offered on Amazon ElasKc Map- Reduce Used by LinkedIn, Neqlix, Salesforce, Twiier, Yahoo... At Yahoo, as of early 2011: 1000s of jobs/day 75%+ of Hadoop jobs Has an interackve example- generator command, but no side- by- side UI L

13 Next: NOVA storage & processing workflow manager e.g. Nova dataflow programming framework e.g. Pig distributed sorkng & hashing e.g. Map- Reduce scalable file system e.g. GFS Debugging aides: Before: example data generator During: instrumentakon framework ANer: provenance metadata manager

14 Why a Workflow Manager? ConKnuous data processing (simulated on top of Pig/Hadoop stakc processing layer) Independent scheduling of workflow modules

15 Example RSS feed NEW Workflow template deteckon ALL news arkcles ALL news site templates ALL NEW template tagging NEW NEW shingling NEW shingle hashes seen ALL NEW NEW de- duping NEW unique arkcles

16 Data Passes Through Many Sub- Systems Nova datum X ingeskon Pig Map- Reduce low- latency processor serving datum Y GFS provenance of X? metadata queries

17 Ibis Project metadata Ibis metadata queries integrated metadata answers data processing sub- systems metadata manager users Benefits: Provide uniform view to users Factor out metadata management code Decouple metadata lifekme from data/subsystem lifekme Challenges: Overhead of shipping metadata Disparate data/processing granularikes

18 Example Data and Process GranulariKes Workflow Table MR program Pig script Web page Row Cell Column group Column AlternaKve Pig job MR job MR job phase MR task Pig logical operakon Pig physical operakon Task aiempt data granulari7es process granulari7es

19 What s Hard About MulK- Granularity Provenance? Inference: Given relakonships expressed at one granularity, answer queries about other granularikes (the seman;cs are tricky here!) Efficiency: Implement inference without resorkng to materializing everything in terms of finest granularity (e.g. cells)

20 Inferring TransiKve Links map phase reduce phase IMDB web page Yahoo! Movies web page IMDB extracted table )tle year lead actor Avatar 2009 Worthington IncepKon 2010 DiCaprio Yahoo extracted table )tle year lead actor Avatar 2009 Saldana IncepKon 2010 DiCaprio map output 1 map output 2 combined table )tle year lead actor Avatar 2009 A1: Worthington A2: Saldana IncepKon 2010 DiCaprio Q: Which web page said Saldana?

21 Next: INSPECTOR GADGET storage & processing workflow manager e.g. Nova dataflow programming framework e.g. Pig distributed sorkng & hashing e.g. Map- Reduce scalable file system e.g. GFS Debugging aides: Before: example data generator During: instrumentakon framework ANer: provenance metadata manager

22 MoKvated by User Interviews Interviewed 10 Yahoo dataflow programmers (mostly Pig users; some users of other dataflow environments) Asked them how they (wish they could) debug

23 Summary of User Interviews # of requests feature 7 crash culprit determinakon 5 row- level integrity alerts 4 table- level integrity alerts 4 data samples 3 data summaries 3 memory use monitoring 3 backward tracing (provenance) 2 forward tracing 2 golden data/logic teskng 2 step- through debugging 2 latency alerts 1 latency profiling 1 overhead profiling 1 trial runs

24 Our Approach Goal: a programming framework for adding these behaviors, and others, to Pig Precept: avoid modifying Pig or tampering with data flowing through Pig Approach: perform Pig script rewrikng insert special UDFs that look like no- ops to Pig

25 Pig w/ Inspector Gadget load filter load join IG coordinator group count store

26 Example: Integrity Alerts load filter load alert! join IG coordinator propagate alert to user group count store

27 Example: Crash Culprit Determina;on Phases 1 to n- 1: record counts Phase n: records load filter load join IG coordinator group Phases 1 to n- 1: maintain count lower bounds Phase n: maintain last- seen records count store

28 Example: Forward Tracing load load filter IG coordinator traced records join group tracing instruckons report traced records to user count store

29 Flow end user result dataflow program + app. parameters applicakon IG driver library launch instrumented dataflow run(s) raw result(s) load load IG coordinator filter join dataflow engine run;me store

30 Agent & Coordinator APIs Agent Class init(args) tags = observerecord(record, tags) receivemessage(source, message) finish() Agent Messaging sendtocoordinator(message) sendtoagent(agentid, message) senddownstream(message) sendupstream(message) Coordinator Class init(args) receivemessage(source, message) output = finish() Coordinator Messaging sendtoagent(agentid, message)

31 ApplicaKons Developed For IG # of requests feature lines of code (Java) 7 crash culprit determinakon row- level integrity alerts 89 4 table- level integrity alerts 99 4 data samples 97 3 data summaries memory use monitoring N/A 3 backward tracing (provenance) forward tracing golden data/logic teskng step- through debugging N/A 2 latency alerts latency profiling overhead profiling trial runs 93

32 RubySky: ScripKng Framework with Built- in Debugging Support Scenario: ad- hoc data analysis e.g. we just did a crawl; is it too biased toward Yahoo content? In this scenario, somekmes using new data; always running new code à Bugs are the norm à Debugging should be a first class cikzen!

33 RubySky Example :Crawl_data <= load(context, "/crawl_data", schema {string :url;...})!! # quick-and-dirty determination of "domain" from "url"! # (e.g. => yahoo.com)! :With_domains <= Crawl_data.foreach() do r! url = r[:url]! prefix = if (url[0, prefix.len] == prefix)! domain = url[prefix.len, url.index('/', prefix.len) - prefix.len]! else! throw Don t know how to parse this url: + url! end! Tuple.new([url, domain])! end!! :Yahoo_only <= With_domains.filter do r! r[:domain] =~ /^yahoo/! end!! :Grouped <= Yahoo_only.group!! :Grouped.each { r puts NUM. YAHOO URLs: + r[:yahoo_only].size }!

34 RubySky Example: w/debug stmts...! :With_domains <= Crawl_data.foreach() do r!...! else! domain = PUNT url! end! Tuple.new([url, domain])! end!! # examine a few extracted domains, as sanity check! :Domain_sample <= With_domains.foreach() do r! $count += 1! ($count < SAMPLE_SIZE)? [Tuple.new([r[:domain]])] : []! end! SHOW Domain_sample!! # alert if any extracted domain is empty or null! :Domain_missing <= With_domains.foreach() do r! (r[:domain] == "" or r[:domain] == nil)?! [Tuple.new(["ALERT! Missing domain from: " + r[:url]])] : []! end! SHOW Domain_missing!!...!

35 RubySky ExecuKon launch load client punt request response (code and/or data) domain samples empty domain alerts extract domain spy agent filter group answer (# yahoo URLs) count

36 Summary storage & processing workflow manager e.g. Nova dataflow programming framework e.g. Pig distributed sorkng & hashing e.g. Map- Reduce scalable file system e.g. GFS Nova Pig Debugging aides: Before: example data generator During: instrumentakon framework ANer: provenance metadata manager Dataflow Illustrator Inspector Gadget, RubySky Ibis

37 Related Work Pig: DryadLINQ, Hive, Jaql, Scope, rela;onal query languages Nova: BigTable, CBP, Oozie, Percolator, scien;fic workflow, incremental view maintenance Dataflow illustrator: [Mannila/Raiha, PODS 86], reverse query processing, constraint databases, hardware verifica;on & model checking Inspector gadget, RubySky: XTrace, taint tracking, aspect- oriented programming Ibis: Kepler COMAD, ZOOM user views, provenance management for databases & scien;fic workflows

38 Collaborators Shubham Chopra Tyson Condie Anish Das Sarma Alan Gates Pradeep Kamath Ravi Kumar Shravan Narayanamurthy Olga Natkovich Benjamin Reed Santhosh Srinivasan Utkarsh Srivastava Andrew Tomkins