2. Cascading and Windows Azure HDInsight
First, some highly recommended reading…
Using MapReduce with HDInsight
http://www.windowsazure.com/en-us/manage/
services/hdinsight/using-mapreduce-with-
hdinsight/
Drive Smarter Decisions with Hadoop and
Windows Azure HDInsight
http://www.slideshare.net/Hadoop_Summit/
drive-smarter-decisions-with-hadoop-and-
windows-azure-hdinsight
2
3. Cascading and Windows Azure HDInsight
Second, a highly recommended talk…
Hadoop Summit 2013 San Jose
Wed, June 26th, 2:05-2:45 pm
hadoopsummit.org/san-jose/schedule/
“Building Tools for the Hadoop Developer”
In this session we’ll first discuss our experience extending Hadoop development to
new platforms & languages and then discuss our experiments and experiences
building supporting developer tools and plugins for those platforms. First, we’ll take a
hands on approach to showing our experiments and successes extending Hadoop to
languages such as JavaScript and .NET with LINQ. Second, we’ll walk through some
of the developer & developer ops tools and plugins we’ve experimented with in an
effort to simplify life for the Hadoop developer across both on premises and cloud-
based projects.
Matt Winkler, Microsoft
blogs.msdn.com/mwinkle
@mwinkle
3
5. Cascading – origins
API author Chris Wensel worked as a system architect
at an Enterprise firm well-known for many popular
data products.
Wensel was following the Nutch open source project –
where Hadoop started.
Observation: would be difficult to find Java developers
to write complex Enterprise apps in MapReduce –
potential blocker for leveraging new open source
technology.
5
6. Cascading – functional programming
Key insight: MapReduce is based on functional programming
– back to LISP in 1970s. Apache Hadoop use cases are
mostly about data pipelines, which are functional in nature.
To ease staffing problems as “Main Street” Enterprise firms
began to embrace Hadoop, Cascading was introduced
in late 2007, as a new Java API to implement functional
programming for large-scale data workflows:
• leverages JVM and Java-based tools without any
need to create new languages
• allows programmers who have J2EE expertise
to leverage the economics of Hadoop clusters
6
13. void map (String doc_id, String text):
for each word w in segment(text):
emit(w, "1");
void reduce (String word, Iterator group):
int count = 0;
for each pc in group:
count += Int(pc);
emit(word, String(count));
The Ubiquitous Word Count
Definition:
count how often each word appears
in a collection of text documents
This simple program provides an excellent test case for
parallel processing, since it illustrates:
• requires a minimal amount of code
• demonstrates use of both symbolic and numeric values
• shows a dependency graph of tuples as an abstraction
• is not many steps away from useful search indexing
• serves as a “HelloWorld” for Hadoop apps
Any distributed computing framework which can runWord
Count efficiently in parallel at scale can handle much
larger and more interesting compute problems.
Document
Collection
Word
Count
Tokenize
GroupBy
token Count
R
M
count how often each word appears
in a collection of text documents
13
17. (ns impatient.core
(:use [cascalog.api]
[cascalog.more-taps :only (hfs-delimited)])
(:require [clojure.string :as s]
[cascalog.ops :as c])
(:gen-class))
(defmapcatop split [line]
"reads in a line of string and splits it by regex"
(s/split line #"[[](),.)s]+"))
(defn -main [in out & args]
(?<- (hfs-delimited out)
[?word ?count]
((hfs-delimited in :skip-header? true) _ ?line)
(split ?line :> ?word)
(c/count ?count)))
; Paul Lam
; github.com/Quantisan/Impatient
word count – Cascalog / Clojure
Document
Collection
Word
Count
Tokenize
GroupBy
token Count
R
M
17
18. github.com/nathanmarz/cascalog/wiki
• implements Datalog in Clojure, with predicates backed
by Cascading – for a highly declarative language
• run ad-hoc queries from the Clojure REPL –
approx. 10:1 code reduction compared with SQL
• composable subqueries, used for test-driven development
(TDD) practices at scale
• Leiningen build: simple, no surprises, in Clojure itself
• more new deployments than other Cascading DSLs –
Climate Corp is largest use case: 90% Clojure/Cascalog
• has a learning curve, limited number of Clojure developers
• aggregators are the magic, and those take effort to learn
word count – Cascalog / Clojure
Document
Collection
Word
Count
Tokenize
GroupBy
token Count
R
M
18
19. import com.twitter.scalding._
class WordCount(args : Args) extends Job(args) {
Tsv(args("doc"),
('doc_id, 'text),
skipHeader = true)
.read
.flatMap('text -> 'token) {
text : String => text.split("[ [](),.]")
}
.groupBy('token) { _.size('count) }
.write(Tsv(args("wc"), writeHeader = true))
}
word count – Scalding / Scala
Document
Collection
Word
Count
Tokenize
GroupBy
token Count
R
M
19
20. github.com/twitter/scalding/wiki
• extends the Scala collections API so that distributed lists
become “pipes” backed by Cascading
• code is compact, easy to understand
• nearly 1:1 between elements of conceptual flow diagram
and function calls
• extensive libraries are available for linear algebra, abstract
algebra, machine learning – e.g., Matrix API, Algebird, etc.
• significant investments by Twitter, Etsy, eBay, etc.
• great for data services at scale
• less learning curve than Cascalog
word count – Scalding / Scala
Document
Collection
Word
Count
Tokenize
GroupBy
token Count
R
M
20
21. github.com/twitter/scalding/wiki
• extends the Scala collections API so that distributed lists
become “pipes” backed by Cascading
• code is compact, easy to understand
• nearly 1:1 between elements of conceptual flow diagram
and function calls
• extensive libraries are available for linear algebra, abstract
algebra, machine learning – e.g., Matrix API,Algebird, etc.
• significant investments by Twitter, Etsy, eBay, etc.
• great for data services at scale
• less learning curve than Cascalog
word count – Scalding / Scala
Document
Collection
Word
Count
Tokenize
GroupBy
token Count
R
M
Cascalog and Scalding DSLs
leverage the functional aspects
of MapReduce, helping limit
complexity in process
21
31. Cascading – functional programming
• Twitter, eBay, LinkedIn, Nokia, YieldBot, uSwitch, etc.,
have invested in open source projects atop Cascading
– used for their large-scale production deployments
• new case studies for Cascading apps are mostly
based on domain-specific languages (DSLs) in JVM
languages which emphasize functional programming:
Cascalog in Clojure (2010)
Scalding in Scala (2012)
github.com/nathanmarz/cascalog/wiki
github.com/twitter/scalding/wiki
Why Adopting the Declarative Programming PracticesWill ImproveYour Return fromTechnology
Dan Woods, 2013-04-17 Forbes
forbes.com/sites/danwoods/2013/04/17/why-adopting-the-declarative-programming-
practices-will-improve-your-return-from-technology/
31
32. Functional Programming for Big Data
WordCount with token scrubbing…
Apache Hive: 52 lines HQL + 8 lines Python (UDF)
compared to
Scalding: 18 lines Scala/Cascading
functional programming languages help reduce
software engineering costs at scale, over time
32
33. Workflow Abstraction – pattern language
Cascading uses a “plumbing” metaphor in the Java API,
to define workflows out of familiar elements: Pipes, Taps,
Tuple Flows, Filters, Joins, Traps, etc.
Scrub
token
Document
Collection
Tokenize
Word
Count
GroupBy
token
Count
Stop Word
List
Regex
token
HashJoin
Left
RHS
M
R
Data is represented as flows of tuples. Operations within
the flows bring functional programming aspects into Java
In formal terms, this provides a pattern language
33
34. Pattern Language
structured method for solving large, complex design
problems, where the syntax of the language ensures
the use of best practices – i.e., conveying expertise
Failure
Traps
bonus
allocation
employee
PMML
classifier
quarterly
sales
Join
Count
leads
A Pattern Language
Christopher Alexander, et al.
amazon.com/dp/0195019199
34
35. Workflow Abstraction – literate programming
Cascading workflows generate their own visual
documentation: flow diagrams
in formal terms, flow diagrams leverage a methodology
called literate programming
provides intuitive, visual representations for apps –
great for cross-team collaboration
Scrub
token
Document
Collection
Tokenize
Word
Count
GroupBy
token
Count
Stop Word
List
Regex
token
HashJoin
Left
RHS
M
R
35
36. references…
by Don Knuth
Literate Programming
Univ of Chicago Press, 1992
literateprogramming.com/
“Instead of imagining that our main task is
to instruct a computer what to do, let us
concentrate rather on explaining to human
beings what we want a computer to do.”
36
37. Workflow Abstraction – business process
following the essence of literate programming, Cascading
workflows provide statements of business process
this recalls a sense of business process management
for Enterprise apps (think BPM/BPEL for Big Data)
Cascading creates a separation of concerns between
business process and implementation details (Hadoop, etc.)
this is especially apparent in large-scale Cascalog apps:
“Specify what you require, not how to achieve it.”
by virtue of the pattern language, the flow planner then
determines how to translate business process into efficient,
parallel jobs at scale
37
38. references…
by Edgar Codd
“A relational model of data for large shared data banks”
Communications of the ACM, 1970
dl.acm.org/citation.cfm?id=362685
rather than arguing between SQL vs. NoSQL…
structured vs. unstructured data frameworks…
this approach focuses on what apps do:
the process of structuring data
38
39. Two Avenues to the App Layer…
scale ➞
complexity➞
Enterprise: must contend with
complexity at scale everyday…
incumbents extend current practices and
infrastructure investments – using J2EE,
ANSI SQL, SAS, etc. – to migrate
workflows onto Apache Hadoop while
leveraging existing staff
Start-ups: crave complexity and
scale to become viable…
new ventures move into Enterprise space
to compete using relatively lean staff,
while leveraging sophisticated engineering
practices, e.g., Cascalog and Scalding
39
45. abstraction RDBMS JVM Cluster
parser ANSI SQL
compliant parser
ANSI SQL
compliant parser
optimizer logical plan,
optimized based on stats
logical plan,
optimized based on stats
planner physical plan API “plumbing”
machine
data
query history,
table stats
app history,
tuple stats
topology b-trees, etc. heterogenous, distributed:
Hadoop, in-memory, etc.
visualization ERD flow diagram
schema table schema tuple schema
catalog relational catalog tap usage DB
provenance (manual audit) data set
producers/consumers
abstraction layers in queries…
45
46. Lingual – articles
Open Source 'Lingual' Helps SQL Devs Unlock Hadoop
Thor Olavsrud, 2013-02-22
cio.com/article/729283/Open_Source_Lingual_Helps_SQL_Devs_Unlock_Hadoop
Hadoop AppsWithout MapReduce Mindsets
Adrian Bridgwater, 2013-02-28
drdobbs.com/open-source/hadoop-apps-without-mapreduce-mindsets/240149708
Concurrent gives old SQL users new Hadoop tricks
Jack Clark, 2013-02-20
theregister.co.uk/2013/02/20/hadoop_sql_translator_lingual_launches/
Concurrent Raises $4 Million,Will Expand Big Data App Framework
James Johnson, 2013-03-20
inquisitr.com/581755/concurrent-raises-4-million-will-expand-big-data-app-framework
Concurrent Releases Lingual, a SQL DSL for Hadoop
Boris Lublinsky, 2013-02-28
infoq.com/news/2013/02/Lingual
46
47. Lingual – JDBC driver
public void run() throws ClassNotFoundException, SQLException {
Class.forName( "cascading.lingual.jdbc.Driver" );
Connection connection =
DriverManager.getConnection(
"jdbc:lingual:local;schemas=src/main/resources/data/example" );
Statement statement = connection.createStatement();
ResultSet resultSet = statement.executeQuery(
"select *n"
+ "from "EXAMPLE"."SALES_FACT_1997" as sn"
+ "join "EXAMPLE"."EMPLOYEE" as en"
+ "on e."EMPID" = s."CUST_ID"" );
while( resultSet.next() ) {
int n = resultSet.getMetaData().getColumnCount();
StringBuilder builder = new StringBuilder();
for( int i = 1; i <= n; i++ ) {
builder.append( ( i > 1 ? "; " : "" )
+ resultSet.getMetaData().getColumnLabel( i )
+ "="
+ resultSet.getObject( i ) );
}
System.out.println( builder );
}
resultSet.close();
statement.close();
connection.close();
}
47
48. Lingual – JDBC result set
$ gradle clean jar
$ hadoop jar build/libs/lingual-examples–1.0.0-wip-dev.jar
CUST_ID=100; PROD_ID=10; EMPID=100; NAME=Bill
CUST_ID=150; PROD_ID=20; EMPID=150; NAME=Sebastian
Caveat: if you absolutely positively must have sub-second
SQL query response for Pb-scale data on a 1000+ node
cluster… Good luck with that! (call the MPP vendors)
This ANSI SQL library is primarily intended for batch
workflows – high throughput, not low-latency –
for many under-represented use cases in Enterprise IT.
In other words, SQL as a DSL.
cascading.org/lingual
48
49. # load the JDBC package
library(RJDBC)
# set up the driver
drv <- JDBC("cascading.lingual.jdbc.Driver",
"~/src/concur/lingual/lingual-local/build/libs/lingual-local-1.0.0-wip-dev-jdbc.jar")
# set up a database connection to a local repository
connection <- dbConnect(drv,
"jdbc:lingual:local;catalog=~/src/concur/lingual/lingual-examples/
tables;schema=EMPLOYEES")
# query the repository: in this case the MySQL sample database (CSV files)
df <- dbGetQuery(connection,
"SELECT * FROM EMPLOYEES.EMPLOYEES WHERE FIRST_NAME = 'Gina'")
head(df)
# use R functions to summarize and visualize part of the data
df$hire_age <- as.integer(as.Date(df$HIRE_DATE) - as.Date(df$BIRTH_DATE)) / 365.25
summary(df$hire_age)
library(ggplot2)
m <- ggplot(df, aes(x=hire_age))
m <- m + ggtitle("Age at hire, people named Gina")
m + geom_histogram(binwidth=1, aes(y=..density.., fill=..count..)) + geom_density()
Lingual – connecting Hadoop and R
49
50. > summary(df$hire_age)
Min. 1st Qu. Median Mean 3rd Qu. Max.
20.86 27.89 31.70 31.61 35.01 43.92
Lingual – connecting Hadoop and R
cascading.org/lingual
50
53. ## train a RandomForest model
f <- as.formula("as.factor(label) ~ .")
fit <- randomForest(f, data_train, ntree=50)
## test the model on the holdout test set
print(fit$importance)
print(fit)
predicted <- predict(fit, data)
data$predicted <- predicted
confuse <- table(pred = predicted, true = data[,1])
print(confuse)
## export predicted labels to TSV
write.table(data, file=paste(dat_folder, "sample.tsv", sep="/"),
quote=FALSE, sep="t", row.names=FALSE)
## export RF model to PMML
saveXML(pmml(fit), file=paste(dat_folder, "sample.rf.xml", sep="/"))
Pattern – create a model in R
53
57. ## run an RF classifier at scale
hadoop jar build/libs/pattern.jar data/sample.tsv out/classify out/trap
--pmml data/sample.rf.xml
## run an RF classifier at scale, assert regression test, measure confusion matrix
hadoop jar build/libs/pattern.jar data/sample.tsv out/classify out/trap
--pmml data/sample.rf.xml --assert --measure out/measure
## run a predictive model at scale, measure RMSE
hadoop jar build/libs/pattern.jar data/iris.lm_p.tsv out/classify out/trap
--pmml data/iris.lm_p.xml --rmse out/measure
Pattern – score a model, using pre-defined Cascading app
57
58. • established XML standard for predictive model markup
• organized by Data Mining Group (DMG), since 1997
http://dmg.org/
• members: IBM, SAS, Visa, NASA, Equifax, Microstrategy,
Microsoft, etc.
• PMML concepts for metadata, ensembles, etc., translate
directly into Cascading tuple flows
“PMML is the leading standard for statistical and data mining models and
supported by over 20 vendors and organizations.With PMML, it is easy
to develop a model on one system using one application and deploy the
model on another system using another application.”
PMML – standard
wikipedia.org/wiki/Predictive_Model_Markup_Language
58
59. • Association Rules: AssociationModel element
• Cluster Models: ClusteringModel element
• Decision Trees: TreeModel element
• Naïve Bayes Classifiers: NaiveBayesModel element
• Neural Networks: NeuralNetwork element
• Regression: RegressionModel and GeneralRegressionModel elements
• Rulesets: RuleSetModel element
• Sequences: SequenceModel element
• SupportVector Machines: SupportVectorMachineModel element
• Text Models: TextModel element
• Time Series: TimeSeriesModel element
PMML – model coverage
ibm.com/developerworks/industry/library/ind-PMML2/
59
61. roadmap – existing algorithms for scoring
•
Random Forest
• Decision Trees
• Linear Regression
• GLM
• Logistic Regression
• K-Means Clustering
• Hierarchical Clustering
• SupportVector Machines
• Multinomial
also, model chaining and general support for ensembles
cascading.org/pattern
61
62. roadmap – top priorities for creating models at scale
•
Random Forest
• Logistic Regression
• K-Means Clustering
a wealth of recent research indicates many opportunities
to parallelize popular algorithms for training models at scale
on Apache Hadoop…
cascading.org/pattern
62
63. roadmap – next priorities for scoring
•
Time Series (ARIMA forecast)
• Association Rules (basket analysis)
• Naïve Bayes
• Neural Networks
algorithms extended based on customer use cases –
contact groups.google.com/forum/?fromgroups#!forum/pattern-user
cascading.org/pattern
63
64. experiments – comparing models
• much customer interest in leveraging Cascading and
Apache Hadoop to run customer experiments at scale
• run multiple variants, then measure relative “lift”
• Concurrent runtime – tag and track models
the following example compares two models trained
with different machine learning algorithms
this is exaggerated, one has an important variable
intentionally omitted to help illustrate the experiment
64
65. ## train a Random Forest model
## example: http://mkseo.pe.kr/stats/?p=220
f <- as.formula("as.factor(label) ~ var0 + var1 + var2")
fit <- randomForest(f, data=data, proximity=TRUE, ntree=25)
print(fit)
saveXML(pmml(fit), file=paste(out_folder, "sample.rf.xml", sep="/"))
experiments – Random Forest model
OOB estimate of error rate: 14%
Confusion matrix:
0 1 class.error
0 69 16 0.1882353
1 12 103 0.1043478
65
66. ## train a Logistic Regression model (special case of GLM)
## example: http://www.stat.cmu.edu/~cshalizi/490/clustering/clustering01.r
f <- as.formula("as.factor(label) ~ var0 + var2")
fit <- glm(f, family=binomial, data=data)
print(summary(fit))
saveXML(pmml(fit), file=paste(out_folder, "sample.lr.xml", sep="/"))
experiments – Logistic Regression model
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 1.8524 0.3803 4.871 1.11e-06 ***
var0 -1.3755 0.4355 -3.159 0.00159 **
var2 -3.7742 0.5794 -6.514 7.30e-11 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01
‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
NB: this model has “var1” intentionally omitted
66
67. experiments – comparing results
•
use a confusion matrix to compare results for the classifiers
• Logistic Regression has a lower “false negative” rate (5% vs. 11%)
however it has a much higher “false positive” rate (52% vs. 14%)
• assign a cost model to select a winner –
for example, in an ecommerce anti-fraud classifier:
FN ∼ chargeback risk
FP ∼ customer support costs
67
69. Anatomy of an Enterprise app
Definition a typical Enterprise workflow which crosses through
multiple departments, languages, and technologies…
ETL
data
prep
predictive
model
data
sources
end
uses
69
70. Anatomy of an Enterprise app
Definition a typical Enterprise workflow which crosses through
multiple departments, languages, and technologies…
ETL
data
prep
predictive
model
data
sources
end
uses
ANSI SQL for ETL
70
71. Anatomy of an Enterprise app
Definition a typical Enterprise workflow which crosses through
multiple departments, languages, and technologies…
ETL
data
prep
predictive
model
data
sources
end
usesJ2EE for business logic
71
72. Anatomy of an Enterprise app
Definition a typical Enterprise workflow which crosses through
multiple departments, languages, and technologies…
ETL
data
prep
predictive
model
data
sources
end
uses
SAS for predictive models
72
73. Anatomy of an Enterprise app
Definition a typical Enterprise workflow which crosses through
multiple departments, languages, and technologies…
ETL
data
prep
predictive
model
data
sources
end
uses
SAS for predictive modelsANSI SQL for ETL most of the licensing costs…
73
74. Anatomy of an Enterprise app
Definition a typical Enterprise workflow which crosses through
multiple departments, languages, and technologies…
ETL
data
prep
predictive
model
data
sources
end
usesJ2EE for business logic
most of the project costs…
74
75. ETL
data
prep
predictive
model
data
sources
end
uses
Lingual:
DW → ANSI SQL
Pattern:
SAS, R, etc. → PMML
business logic in Java,
Clojure, Scala, etc.
sink taps for
Memcached, HBase,
MongoDB, etc.
source taps for
Cassandra, JDBC,
Splunk, etc.
Anatomy of an Enterprise app
Cascading allows multiple departments to combine their workflow components
into an integrated app – one among many, typically – based on 100% open source
a compiler sees it all…
cascading.org
75
76. a compiler sees it all…
ETL
data
prep
predictive
model
data
sources
end
uses
Lingual:
DW → ANSI SQL
Pattern:
SAS, R, etc. → PMML
business logic in Java,
Clojure, Scala, etc.
sink taps for
Memcached, HBase,
MongoDB, etc.
source taps for
Cassandra, JDBC,
Splunk, etc.
Anatomy of an Enterprise app
Cascading allows multiple departments to combine their workflow components
into an integrated app – one among many, typically – based on 100% open source
FlowDef flowDef = FlowDef.flowDef()
.setName( "etl" )
.addSource( "example.employee", emplTap )
.addSource( "example.sales", salesTap )
.addSink( "results", resultsTap );
SQLPlanner sqlPlanner = new SQLPlanner()
.setSql( sqlStatement );
flowDef.addAssemblyPlanner( sqlPlanner );
cascading.org
76
77. a compiler sees it all…
ETL
data
prep
predictive
model
data
sources
end
uses
Lingual:
DW → ANSI SQL
Pattern:
SAS, R, etc. → PMML
business logic in Java,
Clojure, Scala, etc.
sink taps for
Memcached, HBase,
MongoDB, etc.
source taps for
Cassandra, JDBC,
Splunk, etc.
Anatomy of an Enterprise app
Cascading allows multiple departments to combine their workflow components
into an integrated app – one among many, typically – based on 100% open source
FlowDef flowDef = FlowDef.flowDef()
.setName( "classifier" )
.addSource( "input", inputTap )
.addSink( "classify", classifyTap );
PMMLPlanner pmmlPlanner = new PMMLPlanner()
.setPMMLInput( new File( pmmlModel ) )
.retainOnlyActiveIncomingFields();
flowDef.addAssemblyPlanner( pmmlPlanner );
77
78. cascading.org
ETL
data
prep
predictive
model
data
sources
end
uses
Lingual:
DW → ANSI SQL
Pattern:
SAS, R, etc. → PMML
business logic in Java,
Clojure, Scala, etc.
sink taps for
Memcached, HBase,
MongoDB, etc.
source taps for
Cassandra, JDBC,
Splunk, etc.
Anatomy of an Enterprise app
Cascading allows multiple departments to combine their workflow components
into an integrated app – one among many, typically – based on 100% open source
visual collaboration for the business logic is a great
way to improve how teams work together
Failure
Traps
bonus
allocation
employee
PMML
classifier
quarterly
sales
Join
Count
leads
78
79. ETL
data
prep
predictive
model
data
sources
end
uses
Lingual:
DW → ANSI SQL
Pattern:
SAS, R, etc. → PMML
business logic in Java,
Clojure, Scala, etc.
sink taps for
Memcached, HBase,
MongoDB, etc.
source taps for
Cassandra, JDBC,
Splunk, etc.
Anatomy of an Enterprise app
Cascading allows multiple departments to combine their workflow components
into an integrated app – one among many, typically – based on 100% open source
Failure
Traps
bonus
allocation
employee
PMML
classifier
quarterly
sales
Join
Count
leads
multiple departments, working in their respective
frameworks, integrate results into a combined app,
which runs at scale on a cluster… business process
combined in a common space (DAG) for flow
planners, compiler, optimization, troubleshooting,
exception handling, notifications, security audit,
performance monitoring, etc.
cascading.org
79