Docs / Strand / advanced/performance

Performance Optimization

Tips for optimizing workflow performance in Strand.

Overview#

Optimize your workflows for better performance, lower latency, and reduced resource usage.

Best Practices#

Node Configuration#

Parallel Execution#

Workflows automatically execute independent nodes in parallel:

        ┌─→ [Process A]
[Start] ├─→ [Process B]  ← These run in parallel
        └─→ [Process C]

Data Size#

Template Performance#

• Direct field access is faster than JSONPath
• Use JSONPath only when needed for complex queries
• Cache computed values when possible

HTTP Requests#

• Use connectors to reuse connections
• Batch requests when possible
• Handle timeouts appropriately
• Use appropriate HTTP methods (GET for reads, POST for writes)

Monitoring#

Execution Time#

Monitor workflow execution times in the Runs page:

• Identify slow nodes
• Optimize bottlenecks
• Track performance over time

Resource Usage#

• Keep an eye on which workflows run most frequently and how much data they process
• Split very high-volume workflows into smaller, focused ones where possible
• Optimize workflows that consume excessive resources (see the strategies below)

Optimization Strategies#

1. Filter Early#

[Event] → [Filter] → [Process]  ← Filter before processing

2. Minimize Transformations#

Only transform data when necessary. Each transform adds processing time.

3. Use Appropriate Node Types#

• Use Filter for conditions (faster than conditional edges)
• Use Transform only when reshaping data
• Use Print sparingly (adds logging overhead)

4. Optimize Templates#

Fast:

{{ payload.user_id }}

Slower (but more flexible):

{{ payload | jsonpath('$.user.id') }}

Use the right tool for the job!

Related#

• Node Types Reference - Choosing the right nodes
• Templating - Efficient templating

← Back to Strand docs