We will discuss the use of FME Server Automations as the solution for a large-scale data migration project for the second generation of the National Childcare Information System - TCM.
TCM marks a significant milestone in child protection for the agency and the country, and allows all 17 social work areas access to one integrated, up to date information system to manage child protection and welfare cases in the state. The unique approach we took was using FME Server Automations, chained together to migrate this national dataset. We used a distributed deployment of FME Server across three servers, supported by 18 FME Server engines.
This use case is non-spatial and showcases how IT departments can leverage FME Server for non-spatial tasks like data quality, migrations and much more. The presentation will provide insight into managing the volume and scale of data, and lessons learned using FME server automations.
5. The
Peak
of
Data
Integration
20
23
Legacy Database (1st Generation)
● National Child Protection Database (NCCIS)
● Significant milestone in Child Protection for the
Country
● All 17 social work areas had access one integrated
system
6. The
Peak
of
Data
Integration
20
23
As business processes evolved, a new system/
database became necessary.
● Outdated Technology
● Based on CRM 2011/SharePoint 2010
● Adhering to International Standards
● Evolving Case Management Needs
11. The
Peak
of
Data
Integration
20
23
Proposed Solution
● Build an FME Form workspace for each DB table
● Orchestrate workspaces in two automations
● Core Entities (Users, Roles, etc.)
● Forms (Referral Form, Intake Form, etc.)
● Generate audit reports
● Summarized HTML report
● Detailed CSV reports (Forms)
13. The
Peak
of
Data
Integration
20
23
Automation Design (Core Entities)
● Manual Trigger
● Triggered many times during testing
● Clean-Up & Set-Up
● Deleting support files created in the previous run
● Recording details about current run
14. The
Peak
of
Data
Integration
20
23
Automation Design (Core Entities)
● Main Part (Sequence)
● Migration of tables that are dependent of
each other
● Example: user has assigned role;
therefore, role must be written first
● Use of parameterized Database connections
21. The
Peak
of
Data
Integration
20
23
Processing Power
● Originally: 1 box with 3 engines
● Upgraded: Distributed install
● 3 Azure boxes
● 6 engines
● 18 engines in total
● Did not pay off immediately!
● Troubleshooting
● FME Flow Analytics
● External analytics tools
22. The
Peak
of
Data
Integration
20
23
Processing Power
● Originally: 1 box with 3 engines
● Upgraded: Distributed install
● 3 boxes
● 6 engines
● 18 engines in total
● Did not pay off immediately!
● Troubleshooting
● FME Server Analytics
● External analytics tools
23. The
Peak
of
Data
Integration
20
23
Processing Power
● Originally: 1 box with 3 engines
● Upgraded: Distributed install
● 3 boxes
● 6 engines
● 18 engines in total
● Did not pay off immediately!
● Troubleshooting
● FME Server Analytics
● External analytics tools
24. The
Peak
of
Data
Integration
20
23
Other Challenges
● Forms workspaces/process
● Source database complexity (key-value pairs)
● “Unlimited” CPU resources vs. Limited manpower
● Familiarization with business process
● Submitting large number of jobs (FMEFlowJobSubmitter)
● There is a limit
● Not about the number, more about the total size
● Updating workspaces at the same time
● Version control
● Compare Workspaces Tool
● SQL Server writer issue (bulk insert of Core Entities)
● Replaced with JDBC version