UiPath Queue Architecture Fundamentals
Understanding queue design principles enables scalable, resilient automation.
- Orchestrator Queues — distributed transaction store with retry, SLA, and audit capabilities
- Queue item lifecycle — New → In Progress → Successful / BusinessException / ApplicationException
- TransactionItem properties — Reference, SpecificContent (custom data), Priority, DueDate, Output
- Queue configuration — MaxRetryNumber, SLA (minutes), risk SLA warning threshold
- Queue vs non-queue automation — when to use queues (high volume, retry needed) vs simple loops
- Queue item deduplication — using Reference field to prevent duplicate processing
Dispatcher Bot Design
The Dispatcher reads source data and populates the Orchestrator queue for the Performer to process.
- Dispatcher pattern — reads from source (Excel, database, email, API, web scrape) and adds items to queue
- Bulk Add Queue Items — efficient batch upload vs individual Add Queue Item for large datasets
- SpecificContent design — what data to embed in each queue item vs what to re-fetch at process time
- Dispatcher scheduling — time-based triggers, event-driven triggers, API-triggered dispatch
- Dispatcher error handling — partial upload scenarios, duplicate prevention, idempotency
- Multiple dispatchers — fan-in queue patterns for multi-source automation
Performer Bot Design
The Performer processes queue items from Orchestrator using REFramework.
- Performer = REFramework — Get Transaction Item → Process Transaction → SetTransactionStatus
- Parallel performers — running multiple robot instances against the same queue for throughput
- Performer scaling — adding robots dynamically based on queue depth via Orchestrator API
- Process isolation — each transaction item processed independently with full Init recovery on failure
- Application state management between transactions — when to reuse vs restart application sessions
- Performer SLA alignment — respecting DueDate field on queue items for time-sensitive processing
Retry Strategy and Exception Handling
Correct retry design determines bot reliability under real-world failure conditions.
- ApplicationException retry — Orchestrator retries automatically up to MaxRetryNumber
- BusinessException — no retry, item marked failed, logged for human review
- Custom retry logic — implementing additional retry within Process Transaction for specific scenarios
- Retry counter tracking — monitoring retry counts per item to identify systemic failures
- Poison queue patterns — routing permanently failed items to a separate review queue
- Dead letter handling — items that exhaust all retries, escalation workflow design
Throughput, SLA, and Production Monitoring
Queue throughput monitoring and SLA management are essential for enterprise RPA operations.
- Queue throughput metrics — items per hour, processing time per item, failure rate tracking
- SLA monitoring in Orchestrator — risk SLA alerts when items approach due time
- Queue backlog management — identifying and resolving queue pileups before SLA breach
- Queue trigger automation — launching additional performers when queue depth exceeds threshold
- Reporting — Orchestrator queue reports, custom analytics dashboard integration
- Capacity planning — estimating robot count needed for target SLA across queue volumes
Frequently Asked Questions
What queue and dispatcher/performer support do you provide?
We provide real-time support for queue design, dispatcher logic, performer architecture, retry strategy, SLA monitoring, queue backlog resolution, parallel processing design, and production queue failure analysis. We also provide proxy interview support for dispatcher/performer scenario questions.
How do you design dispatcher/performer for high-volume automation?
We help design the full dispatcher/performer architecture — SpecificContent schema for queue items, dispatcher scheduling, bulk queue upload, performer REFramework integration, parallel robot scaling, SLA configuration in Orchestrator, and backlog monitoring. This is one of the most common UiPath interview scenarios and production design challenges we support.
What is the difference between BusinessException and ApplicationException in queues?
BusinessException marks the item as a business failure — no Orchestrator retry, item counted as failed but the bot continues to the next item. ApplicationException triggers Orchestrator retry up to MaxRetryNumber — the performer re-inits and retries the same item. We help implement the correct exception classification logic for your specific business rules.
Can you help with queue backlog and SLA breach situations?
Yes. Queue backlog resolution includes identifying bottlenecks (slow processing, too few robots, performer failures), implementing queue triggers for auto-scaling, adjusting SLA settings, and designing emergency processing workflows. We provide real-time support during production SLA risk situations.
Ready to get real-time expert support?
Same-day start. Confidential. All major time zones covered.