Best practises for IO
- Plan your flow before adding nodes to maintain a clean structure
- Use the search function to quickly find the right node type
- Create flows that have a single, clear purpose, aligned with requirements
- Label flows and variables with consistent names that describe their purpose
- Name nodes descriptively after adding them
- Connect nodes as you go to maintain flow logic visibility
- Test connections as you go to ensure proper data flow
- Use variables instead of hardcoded values
- Design nodes for reuse in library flows
- Combine operations when possible to reduce the node count
- Group related nodes and document their relationships
- Verify node behavior in isolation
- Liberally document what each flow does
- Create new versions before making significant changes
- Minimize crossing lines to keep the flow layout simple
- Validate the input data before processing
- Include error-handling paths
- Ensure that every node output is connected to another node
- Connect success and failure paths correctly
- Test all paths through the flow before activating
- Use node tracking for important flows
- Track integration points
- Use fewer, more capable nodes
- Improve slow-running flows
- Place common conditions first
- Order conditions by likelihood to optimize operations
- Don't repeat operations
- Run independent operations concurrently
- Use batch operations to reduce API calls
- Store frequently accessed data
- Update flows based on usage patterns
- Monitor flow duration, time per node, success/failure rates, variable values at each step, and user abandonment
- Configure session timeout based on flow complexity
- Divide complex workflows into smaller, reusable library flows
- Keep change logs to document the details about each version
- Review flow logs and error messages regularly
- Archive, remove, or deactivate obsolete flows and versions
