Glassbeam: Ad-hoc Analytics on Internet of Complex Things with Apache Cassandra and Apache Spark

© Copyright 2015 Glassbeam Inc.
Ad Hoc Analytics
on
Internet of Complex Things
with
Spark and Cassandra
Mohammed Guller
September 2015

Let’s Take a Quick Poll
 Familiar with IoT
 Data modelling experience in C*
 Familiar with Spark
 Hands-on experience with Spark
3

About Me
 Principal Architect at Glassbeam
 Author of an upcoming book
– “Big Data Analytics with Spark”
 Founded two startups
 Passionate about building new products, big data analytics,
and Machine Learning
 Berkeley Graduate
LinkedIn: www.linkedin.com/in/mohammedguller
Twitter: @MohammedGuller
4

Internet of Things (IoT)
5
Network of objects embedded with software for
collecting and exchanging data over the Internet

Internet of Complex Things (IoCT)
6
 Data Center Devices
– Server, storage, controller
 Medical Devices
– X-Ray, MRI scan, CT scan
 Manufacturing Systems
 Cars
 Electric Vehicle Chargers
 Other Complex Devices
Glassbeam target market is focused on driving opera onal & business
naly cs value for connected product companies in Industrial IoT market
IT & Networks Medical & Health Care EV Chargers & Smart Grid

IT & Networks
Medical &
Healthcare
EV Chargers &
Smart Grid
Industrial & Mfg
Transportation
Glassbeam
7
target market is focused on driving opera onal & business
ue for connected product companies in Industrial IoT market
rks Medical & Health Care
Transporta on
EV Chargers & Smart Grid
Industrial & Mfg
5
Glassbeam target market is focused on d
analy cs value for connected product com
IT & Networks Medical & Health Care
TrIndustrial & Mfg
market is focused on driving opera onal & business
connected product companies in Industrial IoT market
Medical & Health Care EV Chargers & Smart Grid
Transporta on
5
Advanced and
Predictive Analytics
for Connected
Product Companies

10101000101011010101110101111010101000101001010101010111110001011001000110000110101110100110011111000000101
01101010011111000101001010110010100101100010011010101140101010000101010000101111001001101011010010101000001
11101010101000101011010101110101111010101000101001010101010111110001011001000110000110101110100110011111000
00010101101010011111000101001010110010100101100010011010101140101010000101010000101111001001101011010010101
00000100100110101101001001001101011010010010011010001001101011010010010011010110100101101001101001101001101
Analytics on Operational Data
8
Operational Data
to
Powerful Insights

High-level Architecture
9
1010100010101
10101011101011
1101010100010
1001010101010
11111000101100
1000110000110
10111010011001
11110000001010
11010100111110
0010100101011
0010100101100
0100110101011
4010101000010
10100001011110
0100110101101
0010101000001
11101001111001
0011010110100
1010101010100
0101011010101
11010111101010
1000101001010
10101011111000
1011001000110
00011010111010
011
Data
Inges on
Data
Transforma on
Data Stores Middleware Applica ons
Logs
(Streams/
docs)
SPL Library
S
C
A
L
A
R
I
N
F
O
S
E
R
V
E
R
LogVault
Explorer
Workbench
Standard Apps
Custom Apps
Rules & Alerts
DirectAccess
Glassbeam Studio
Cloud Enablement & Automa on
S3 Amazon
Raw logs
Cassandra
Processed Data
Solr Cloud
Index
Analy cs and
Machine learning
Spark
SQL
Spark
Streaming
MLlib
Event Processing & Rules Engine
End to End cloud based architecture built on modern
technologies to handle any machine, any data, any cloud
* SPL (Semiotic Parsing Language) and SCALAR are patent pending technology inventions of Glassbeam

Key Properties of IoCT Data
10
Volume Terabytes of Data
Variety Multi-structured Data
Velocity
Fast Paced Batch Data
Streaming Data

Why We Chose C*
11
Volume Economically Scale from Gigabytes
to Terabytes of Data
Variety Store Multi-structured Data
Velocity
Fast Ingest of New Data Quick
Reload of Old Data
Linear
Scalability
Dynamic
Schema
Fast
Writes

Modeling Data in C*
 Different from Modeling Data in RDBMS
 Queries Drive Table and Primary Key Definitions
– Primary Key Definition Limits the Kind of Queries You Can Run
– C* Does Not Support Joins
12

A Simple Table for Storing Event Data in C*
CREATE TABLE event (
sys_id text,
dt timestamp,
ts timestamp,
severity text,
module text,
message text,
PRIMARY KEY ((sys_id, dt), ts)
) WITH CLUSTERING ORDER BY (ts DESC);
13

Another Table to Filter Events by Severity
CREATE TABLE event_by_severity (
sys_id text,
dt timestamp,
ts timestamp,
severity text,
module text,
message text,
PRIMARY KEY ((sys_id, dt), severity, ts)
) WITH CLUSTERING ORDER BY (severity ASC, ts DESC);
14

Yet Another Table to Filter Events by Module
CREATE TABLE event_by_module (
sys_id text,
dt timestamp,
ts timestamp,
severity text,
module text,
message text,
PRIMARY KEY ((sys_id, dt), module, ts)
) WITH CLUSTERING ORDER BY (module ASC, ts DESC);
15

Ad Hoc Analytics with C*
 Oxymoron
 All queries Must be Known Upfront
16

Workaround Possible but Intractable
Sys_id Model Age OS City State Country
17
• sys_by_model
• sys_by_os
• sys_by_age
• sys_by_state
• sys_by_state_age
• sys_by_age_state
• sys_by_model_age
• sys_by_age_model
• sys_by_age_model_state
• sys_by_model_state_age
• sys_by_model_state_os

Other Barriers to Ad Hoc Queries
 No Aggregation
 No Group By
 No Joins
18

© Copyright 2015 Glassbeam Inc. 19
What Do
I Do
Now?

Spark
21
 Fast and General-purpose Cluster Computing
Framework for Processing Large Datasets
 API in Scala, Java, Python, SQL, and R

Integrated Libraries for a Variety of Tasks
22
Spark Core
Spark
SQL
GraphX
Spark
Streaming
MLlib &
Spark ML

One Minor Problem!
 Spark Does not Have Built-in Support for C*
 Built-in Support for HDFS, S3 and JDBC-compliant
Databases
23

Spark Cassandra Connector
 Open Source Library for Integrating Spark with C*
 Enables a Spark Application to Process Data in C* Just
Like Data from the Built-in Data Sources
24

Spark with C*
 Enables Ad Hoc Analytics
 CQL Limitations No Longer Apply
 Query Data Using SQL/HiveQL
– Filter on Any Column
– Aggregations
– Group By
25

Ad Hoc Analytics in Spark Shell
26

Launch the Spark Shell
/path/to/spark/bin/spark-shell
--master spark://host:7077
--packages com.datastax.spark:spark-cassandra-connector_2.10:1.4.0
27

Create a DataFrame
val events = sqlContext.read
.format("org.apache.spark.sql.cassandra")
.options( Map(
"keyspace" -> "test",
"table" -> "event"))
.load()
28

Fire Queries
events.cache()
events.select("ts", "module", "message").where($"severity" === "ERROR").show
events.select("ts", "severity", "message").where($"module" === "m1").show
events.select("ts", "message").where($"severity" === "ERROR" &&
$"module" === "m1").show
events.groupBy("severity").count()
29

Spark SQL JDBC/ODBC Server
 Analyze data in C* with just SQL/HiveQL
 Command Line Shell
– Beeline
 Graphical SQL Client
– Squirrel
 Data Visualization Applications
– Tableau
– ZoomData
– Qlik
30

Ad hoc Analytics with Spark SQL JDBC/ODBC server
31

Start the Spark SQL JDBC Server
/path/to/spark/sbin/start-thriftserver.sh
--master spark://hostname:7077
--packages com.datastax.spark:spark-cassandra-connector_2.10:1.4.0
32

Launch Beeline From a Terminal
/path/to/spark/bin/beeline
33

Connect to the Spark SQL JDBC Server
beeline> !connect jdbc:hive2://localhost:10000
34

Create a Temporary Table
0: jdbc:hive2://localhost:10000> CREATE TEMPORARY TABLE event
. . . . . . . . . . . . . . . .> USING org.apache.spark.sql.cassandra
. . . . . . . . . . . . . . . .> OPTIONS (
. . . . . . . . . . . . . . . .> keyspace "test",
. . . . . . . . . . . . . . . .> table "event"
. . . . . . . . . . . . . . . .> );
35

Query Data with SQL/HiveQL
...> CACHE TABLE event;
...> SELECT severity, count(1) as total FROM event GROUP BY severity;
...> SELECT module, severity, count(1) FROM event GROUP BY module, severity;
36

Caveats
 Latency
 Spark Query May Require Expensive Table Scan
– Reads Every Row
– Disk I / O Slow
37

Reduce the Impact of Slow Disk I / O
 Cache Tables
 Replace HDD with SSD
 Add More Nodes
38

Recommendations
 Known Queries Requiring Sub-second Response Time
– Query C* Directly
– Create Query Specific Tables
– Pre-aggregate Data
 Ad Hoc Queries
– Spark
39

Glassbeam: Ad-hoc Analytics on Internet of Complex Things with Apache Cassandra and Apache Spark

Recommended

More Related Content

What's hot (19)

Viewers also liked (20)

Similar to Glassbeam: Ad-hoc Analytics on Internet of Complex Things with Apache Cassandra and Apache Spark (20)

More from DataStax Academy (20)

Recently uploaded (20)

Glassbeam: Ad-hoc Analytics on Internet of Complex Things with Apache Cassandra and Apache Spark