IBM Data AI IBMs Complete Test Data Management

IBM Data & AI IBM’s Complete Test Data Management Family IBM Internal and Business Partners Only

IBM’s Complete Test Data Management Family Virtual Test Data Copies Database Subsetting and Masking Synthetic Data Fabrication Data Coverage & Automated API Testing Database subsetting of “complete business object” for lean test data environment Production-like, synthetic test data when production data isn’t available or is too sensitive Complete test data coverage for API testing, automated data assembly, and Dev. Ops integration Info. Sphere Virtual Data P I P E L I N E Lean, near instant database copies with central management and self-service (roll back) refresh 2

IBM’s Complete Test Data Management Family IMS Informix Virtual Full-Copy Production Clone Adabas Info. Sphere Virtual Data P I P E L I N E PRODUCTION GOLD COPY SIT 1 © 2019 IBM Corporation DEV 2 SIT 2 100% Data Coverage & Automated API Testing VSAM Teradata Synthetic Data Fabrication SUBSETTED GOLD COPY INTEGRATED SUBSETS VIRTUAL COPIES DEV 1 Production Data Subset Sybase UAT DEV 1 DEV 2 UAT SIT 1 SIT 2 TRAIN DEV/TEST TRAIN 3

4 Databases Run the Business. . . and are cloned many times 2015 2014 Standby 2013 2012 2011 DR Development PII, PHI, PCI Data 2010 2009 2008 Training PII, PHI, PCI Data 2007 2006 2005 Test PII, PHI, PCI Data

How does test data management help? Requirements Create “right-size” environments with realistic data for application testing and development • • Test Data Management -Subset -Mask • Relational data sets • Benefits -Refresh -Compare 20 TB • Production or Production Clone 20 GB 200 GB Development Integration Test • • Unit Test 100 GB Shorten iterative testing cycles Create referentially intact, “right-sized” test databases Compare data across dev and test iterations Protect confidential data 1 TB UAT • • Speed delivery of new deployments Improve quality by testing more use cases Capture errors earlier in development cycle reducing cost & improving quality Lower cost by reducing storage costs and time Lower risk by protecting sensitive data

Effective Test Data Management Subsets data while preserving referential integrity Producing complete business objects that contain the data and business relationships to accurately test and re-test applications with integrity Retrieval of r eferentially-intact subsets of data across related applications, including the metadata. Customer Management in DB 2 LUW Shipping and Logistics in Oracle Sales Management in DB 2 LUW Federated access to related business objects across the enterprise

Effective Test Data Management Validates and verifies expected and actual results • Verification during Test Data Management projects • Compare Expected Results • Compare Actual Results • Identify Defects • Answers the Questions: • Are data subsets built correctly ? • Is the referential integrity accurate and preserved ? • Was the data masked correctly ? • Why is this Better ? • Compares the actual data values • Flexible capabilities to compare database to database, file to file, data model to data etc. • Can compare multiple data stores • Can compare single and multiple tables simultaneously

IBM Info. Sphere Virtual Data Pipeline (VDP) DEV QA 3 1 Capture App Consistent Data Self Service Virtual VDP 1 TB 2 Copies Store VDP Gold copy on any storage DB 2 INTEGRATION PERF TEST UAT PRE-PROD RECOVERY Instant clone creation 1. 1 TB Storage Used 90% Storage Savings DR

Data Masking Demo © 2019 IBM Corporation

VDP Data Masking Data Refresh happens automatically 6 Ops: Specify data masking routines and data masking server 2 Prototype Configure VDP to create a live clone Unit Test 1 VDP Build Integration MS SQL DB ORACLE DB QA Test Live Clone DB 2 DB 3 VDP mounts the data to data masking server. Data is masked here. 10 UAT STAGING 4 DATA MASKING SERVER After Masking is done, control returns to VDP SUPPORT 5 Testers can select ONLY masked copies