Elevate MongoDB with ODBC/JDBC

Elevate MongoDB with ODBC/JDBC
Sumit Sarkar, Principal Systems Engineer
DataDirect, A Progress Software Company
www.linkedin.com/in/meetsumit
@SAsInSumit

Welcome MongoDB connectivity to the DataDirect family
Big Data/NoSQL
 Apache Hadoop Hive
 Cloudera
 Hortonworks
 Pivotal HD
 MapR
 EMR
 Pivotal HAWQ
 Cloudera Impala
 MongoDB
 Cassandra (preview)
 Spark SQL (preview)
 SAP HANA (ODBC preview)
Data Warehouses
 Amazon Redshift
 SAP Sybase IQ
 Teradata
 Pivotal Greenplum
Relational
 Oracle DB
 Microsoft SQL Server
 IBM DB2
 MySQL
 PostgreSQL
 IBM Informix
 SAP Sybase
 Pervasive SQL
 Progress OpenEdge
 Progress Rollbase
SaaS/Cloud
 Salesforce.com
 Database.com
 FinancialForce
 Veeva CRM
 ServiceMAX
 Any Force.com App
 Hubspot
 Marketo
 Microsoft Dynamics CRM
 Microsoft SQL Azure
 Oracle Eloqua
 Oracle Service Cloud
 Google Analytics
EDI/XML/Text
 EDIFACT
 EDIG@S
 EANCOM
 X12
 IATA
 Healthcare EDI: X12, HIPAA,
ICD-10, HL7
 Custom EDI
 Flat files: CSV, TSV, dBase,
Clipper, Foxpro, Paradox
 Text Files
Any
 SDK
 SequeLink Socket Server
 Customer Engineering

ODBC/JDBC applications with
MongoDB

Why connect to MongoDB with SQL access?
 Analytics and Data Visualization
 Reporting
 Data Virtualization / Federation
 Data Warehousing

Elevate your MongoDB data value with open database standards
Business Intelligence Data Integration
ODBC/JDBC

Why MongoDB Developers care about SQL?
 Your applications are awesome and the data deserves open
access
 Expand footprint of MongoDB where enterprise reporting
requirements prevent adoption
 Share analytics, reporting, integration work with to your
colleagues

Select MongoDB ODBC/JDBC use cases from our user base
 Network Security: Build reports using Microsoft BI Stack for
all incoming network traffic stored in MongoDB
 Enhance Operational Systems: Store order details from IBM
order management system in MongoDB repository that
require SAP reporting
 Visual Analytics: Use Tableau to determine success of
marketing campaign data including clicks, videos, social
shares, etc from MongoLabs
 Complex Analysis: Build cubes for intelligence using
complex MongoDB documents storing clinical trial data

How SQL access to NoSQL works

MongoDB: an OpenSource, NoSQL Document (JSON) Database
MongoDB Data Model
Sample JSON Document:
{ name: “sue”, age: 26, status: “A”, groups: [“news”, “sports”]}
Relational database design
focuses on data storage
NoSQL document database design
focuses on data use

Introduced tools to perfect our intelligent schema
DataDirect ODBC/
JDBC Driver
Meta
data
Schema
Tool
Business Intelligence Data Integration

MongoDB | Schema Design Comparison
Relational Design NoSQL Document Design
{ user: {
first: “Brody,
last: “Messmer”, ...
}
purchases: [
{ symbol: “PRGS”, date: “2013-02-13”, price: 23.50, qty: 100, ...},
{ symbol: “PRGS”, date: “2012-06-12”, price: 20.57, qty: 100, ...},
...
]
}
...
Collection: users
VS
user_id first last …
123456 Brody Messmer …
…
user_id symbol date price qty …
123456 PRGS 2013-02-13 23.50 100 …
123456 PRGS 2012-06-12 20.57 100 …
…
Table: users
Table: purchases

Collection Name: stock
{ symbol: “PRGS”,
purchases:[
{date: ISODate(“2013-02-13T16:58:36Z”), price: 23.50, qty: 100},
{date: ISODate(“2012-06-12T08:00:01Z”), price: 20.57, qty: 100,
sellDate: ISODate(“2013-08-16T12:34:58Z”)}, sellPrice: 24.60}
]
}
MongoDB | Progress DataDirect Approach – Normalize
Table Name: stock
_id symbol
1 PRGS
stock_id Date Price qty sellDate sellPrice
1 2013-02-13
16:58:36
23.50 100 NULL NULL
1 2012-06-12
08:00:01
20.57 100 2013-08-16
12:34:58
24.60
Table Name: stock_purchases

Demo | MongoDB

Tableau and MongoDB ODBC

Lessons Learned

Challenges we faced in building SQL interface to NoSQL
 Mapping SQL to MongoDB query language
• Queries are written as BSON, but exposed in Mongo “drivers” as an API.
• No support for joins
• Lack of Implicit conversions for query filters (where clause)
• Sorting is not ANSI SQL compliant
 Non-relational schema
• Document oriented data model with complex data types (denormalized)
• Self-describing schema -- Can only discover columns by selecting data

SQL to MongoDB Lessons Learned with our users
 Managing logical schema across BI metadata modeling clients and servers/clusters
 Important to consider cross collection joins in stress tests on BI servers
 Different BI tools require specific properties such as max varchar sizes on string data
types, or field names that start with underscore.
 Hard to anticipate document structures and levels of nesting – so we supported
infinitely deep
 Lack of enforced data types from document to document (row to row)
• Driver attempts to convert to the SQL Type defined. If conversion fails, NULL is returned.
• MongoDB does not implicitly convert to/from strings when applying filters (where clause)
– Driver is capable filtering client side via a hidden option
 Field names of different case can result in collisions from document to document
• The driver will expose these fields as separate columns without “uppercaseIdentifiers”=true
 Must sample data to generate schema on read – need to configure this value
“ColumnDiscoverySampleSize”

Elevate MongoDB with ODBC/JDBC

Elevate MongoDB with ODBC/JDBC

Related slideshows

More Related Content

Elevate MongoDB with ODBC/JDBC

Editor's Notes