Caterpillar: Using SnapLogic for Multiple Use Cases

[dmiller]

What were the underlying reasons or business implications for the need to automate business processes?

Many

Describe your automation use case(s) and business process(es).

• Outbound Idoc Traffic: Handling Outbound Idoc traffic from multiple SAP envs presents a challenge in many ways.
  • Our first generation: Our Outbound Idoc Distribution integration automates Idoc data handling using an Idoc Document Listener snap to capture all Outbound Idoc types for a particular SAP connection and route them to IDoc type-specific queue in MQ SERIES. SnapLogic integrations that process IDoc type-specific data read in the data from the queues and process it to various target systems.
  • A second generation version is coming soon that will run on our new Groundplex and use NAS storage in place of MQ.
• Error Handling: I built an Error Pipeline that have been widely adopted in our ORG stack across 200+ project folders. The pipeline offers ServiceNow INC ticket creation and email functionality. Because
  pipelines that reference an error handler do not fail on errors our external scheduling tool
  (Tidal) is removed from the Support model.
• Usage tracking/Customer Charges: I built the first generation of our SnapLogic Usage integration, which read execution log data from the SL public run time API. As we grew, a rebuild was performed changing the data pull from an every hour full pull to a once a day read that scans recursively across all project folders using pagination in the REST snap. This latest version has proven sturdy in handling our rapidly growing execution load.
• Tidal Automation: I built the first generation of our SnapLogic Tidal Automation Model. This model utilizes a Web Service Adapter in Tidal to make API calls to SnapLogic via Datapower using OAuth2 Authentication. This is a self-contained, reliable, model, but does have a max run time of 1 hour due to the 1 hour cloud FW timeout. A second generation model is about to start development that will use a Tidal agent instead of a Web Service Adapter and will include a perl script. The script will handle OAuth2 and loop through execution status checks against the SL public run time API.
• Initialization Errors: Built integration that generates tickets for pipelines that fail though they references an error pipeline. These types of failures fall into a category of infrastructure or initialization issues. This pipeline closes a glaring hole in the support model as node level events or pipeline initialization errors result in error pipelines also failing. This means integrations fail without generating ticket and/or emails