Best Practices: Node Configuration

Overview

As you configure applications in Unqork, you'll notice that some build methods work better than others.

To set you up for success, here are some best practices.

Node Naming

  • Keep your Label and Path names unique for each node in your workflow.

  • Develop a node-naming convention and use it throughout your building process.

  • Use descriptive Labels rather than generic ones. It's easier for your team to know what the nodes are doing with descriptive Labels identifying the node's purpose.

  • Keep Label and Path names short for a cleaner display and smaller JSON payloads.

  • Always label nodes.

General Configuration Best Practices

Build automated lane tasks to complete operations in five minutes to prevent timeouts. User input modules can complete operations as long as needed. The five-minute limitation does not apply to QA or Staging environments. However, workflows promoted to the Production environment are affected by the five-minute operation limit.

  • Create modules in the Automated Workflow lane with the following settings:

    • Enable Server-side Execution Only in module settings for external API service endpoints that are configured in server administration. SSE prevents external services from seeing sensitive information such as financial data or PII (Personal Identifiable Information).

    • Enable Customize RBAC for this Module in the module permissions.

Node Configuration

Below is a glossary of nodes and best practices when using them.

Message Node

Amazon SES Service Limitations:

  • If sending more than 3,000 emails in a single day, inform your CSM (customer success manager) of this requirement.

  • If sending more than 6,000 emails in a single day, discuss best practices with your CSM.

Start, Signal Start, and Timer Start Node

  • Build workflows to complete operations in five minutes to prevent timeouts. The five-minute limitation doesn't apply to QA or Staging environments. However, workflows promoted to the Production environment are affected by the five-minute operation limit.

  • Do not use Start, Signal Start, or Timer Start nodes to initiate nodes that are in a Parallel Gateway process.

Gateway Nodes

Gateway Nodes refer to Exclusive, Inclusive, and Parallel Gateway nodes.

  • Paths from Gateway nodes cannot connect to End, Handoff, or Gateway Split nodes.

  • Configure Gateway Nodes to have at least two paths. A validation error occurs if a Gateway node has only one path.

  • Nodes outside of a Gateway process cannot connect to nodes inside of a Gateway process.

Exclusive Gateway Node

  • Exclusive Gateway Node Links must have a logic formula unless they are the default link. There can only be one default link for each Exclusive Gateway node.

Inclusive Gateway Split Node

  • Build Inclusive Gateway Node Links to use a logic formula for each link path.

Inclusive/Parallel Gateway Split and Join Node Behavior

  • You must have an equal number of Gateway Split and Gateway Join nodes.

  • All Gateway Split nodes must end at a Gateway Join node.

  • All output paths from a Gateway Split node must reach the same Gateway Join Node

Timed Event Node

  • Timed Event nodes can only be added to Task and Sub-Process nodes.

  • The Duration of Timer must be greater than or equal to the durations set for the node's messages. Once the workflow moves onto a new task, the Timed Event is no longer valid and is unable to send the messages.

  • You cannot add a Timed Event node to the Automated swimlane. But, it can trigger a node that exists in the Automated swimlane.

  • You can add Timed Event nodes onto a Sub-Process node. The countdown timer starts if the optional condition is true and when the workflow navigates to any node in the Sub-Process node.

  • To a trigger event to occur, the submission data must be available to the Task or Sub-Process node where the Timed Event node is connected.