High Availability for RPA

High Availability for Appian RPA (RPA HA) helps ensure your robotic process automations continue running reliably, even during infrastructure failures, or other system disruptions. When you enable Appian Cloud HA, Appian automatically includes RPA HA as part of your platform configuration.

Key benefits

RPA HA provides several advantages for your automation infrastructure:

• Continuous Availability: RPA HA keeps all server nodes active and handling traffic simultaneously. If a server node fails, the remaining nodes continue processing requests without interruption, so executions keep running.
• Load Balancing: RPA HA architecture distributes workload across multiple servers to provide consistent performance during peak usage.
• Continuous Monitoring: RPA HA provides real-time health checks across all components, automatically detecting and isolating failed services before they impact your automation workflows.

Key components

Understanding which components the RPA HA infrastructure includes helps you plan your automation environment effectively.

Server-side components

RPA HA protects all server-side components that run in the Appian Cloud infrastructure:

• Operations console and RPA console: The web interface where you manage robots, robot pools, and robotic tasks.
• Orchestration services: The coordination engine that assigns robotic tasks to robots and tracks execution progress.
• Artifact storage: Robotic task packages and dependencies required for deployment and execution.
• Execution history: All execution logs and audit trails.

These components run across three availability zones with automatic failover to ensure continuous operation.

Client-side components

The components that run on your organization's infrastructure are not part of the Appian Cloud HA infrastructure. These are the same components used in a single-node RPA environment and are not changed for RPA HA:

• RPA agents: The software that executes robotic tasks on your host machines.
• Autologin services: Windows services that manage RDP sessions for unattended automation.
• Host machines: The physical or virtual machines where RPA agents run.

Regional availability

RPA HA is available in all Appian Cloud regions where Appian HA is supported. For a full list of supported regions, see Appian Cloud Availability. If Appian HA is available in a region, RPA HA will be as well.

The specific availability zones and disaster recovery capabilities depend on your region:

• Multi-zone regions: RPA HA operates across three availability zones within your Appian Cloud region for maximum resilience.
• Enhanced Business Continuity (EBC): Available in applicable regions that support EBC for cross-region disaster recovery.

Architecture

RPA HA uses a distributed architecture where three RPA console servers operate simultaneously across separate availability zones within your Appian Cloud region.

RPA HA ensures your console remains accessible and tasks continue uninterrupted. This provides protection against zone outages and supports Enhanced Business Continuity (EBC) in applicable regions. For disaster recovery across regions, protection is determined by your Appian environment's EBC configuration.

How it works

RPA HA automatically manages robotic task executions across your available resources:

• Load balancing: A load balancer automatically distributes incoming requests across all healthy servers. When you access the RPA console or when an RPA agent connects to execute tasks, the load balancer routes the request to an available server. If a server becomes unhealthy, the load balancer immediately stops routing traffic to it and directs all requests to the remaining healthy servers.
• Shared state: All servers share access to the same data stores (the MariaDB database and the Redis cache), ensuring consistency across your RPA environment. When you create a robot, deploy a robotic task, or start an execution on one server, that information is immediately available to all other servers. This data consistency ensures that RPA agents can access the same information regardless of which server they connect to.
• RPA agent reconnection: The orchestration system automatically adapts RPA agent connections to the RPA HA environment. If the current server node becomes unavailable, the load balancer routes subsequent requests to healthy nodes. Executions continue running without interruption.

Failover and recovery

RPA HA automatically handles different types of failures to keep your automation running. If a server node or availability zone fails, the load balancer directs all requests to the remaining healthy servers. In-progress executions continue running on a healthy node. Appian RPA continues operating normally while it automatically replaces the failed components.

Enable RPA HA

RPA HA is available to Appian Cloud customers.

Enabling RPA HA involves converting your existing single-server RPA environment to a distributed, multi-zone architecture. The conversion process requires coordination with Appian support.

Prerequisite

Before enabling RPA HA for your environment, ensure you have purchased a Professional or Signature Success Plan (including High Availability).

Request an RPA HA conversion

To enable RPA HA for your environment, open a support case requesting RPA HA configuration for your Appian Cloud environment.

Removed features

Because certain features were originally designed for single-server deployments and are now incompatible with a distributed architecture, Appian has removed them.

RPA queues and scheduling

The RPA queues and RPA scheduling capabilities previously found in the RPA Console were deprecated in RPA version 9.17 and are now fully removed.

REST API endpoints

The following REST API endpoints are fully removed with the release of RPA HA; so, you must update them to use process models and robot pools before upgrading to RPA 9.22. Contact Appian support if you need assistance.

Queue management

• POST /api/v1/queues
• GET /api/v1/queues
• GET /api/v1/queues/{queueId}
• PUT /api/v1/queues/{queueId}
• DELETE /api/v1/queues/{queueId}
• POST /api/v1/queues/{queueId}/items
• GET /api/v1/queues/{queueId}/items
• DELETE /api/v1/queues/{queueId}/items/{itemId}

Schedule management

• POST /api/v1/schedules
• GET /api/v1/schedules
• GET /api/v1/schedules/{scheduleId}
• PUT /api/v1/schedules/{scheduleId}
• DELETE /api/v1/schedules/{scheduleId}
• POST /api/v1/schedules/{scheduleId}/enable
• POST /api/v1/schedules/{scheduleId}/disable

Troubleshooting

If you notice performance degradation after enabling RPA HA, open a support case immediately. Performance issues can indicate configuration problems or resource constraints that require immediate attention from Appian support.

If you suspect one of the three servers is unhealthy but the system hasn't automatically failed over, open a support case immediately. Include details about the suspected issues, error messages, and any performance anomalies you've observed.

Additional resources

For more information about High Availability in Appian:

• High Availability for Appian Cloud
• Appian RPA Console

Feedback