- Getting started
- Best practices
- Tenant
- About the Tenant Context
- Searching for Resources in a Tenant
- Managing Robots
- Connecting Robots to Orchestrator
- Storing Robot Credentials in CyberArk
- Storing Unattended Robot Passwords in Azure Key Vault (read only)
- Storing Unattended Robot Credentials in HashiCorp Vault (read only)
- Storing Unattended Robot Credentials in AWS Secrets Manager (read only)
- Deleting Disconnected and Unresponsive Unattended Sessions
- Robot Authentication
- Robot Authentication With Client Credentials
- Configuring automation capabilities
- Audit
- Settings
- Cloud robots
- Folders Context
- Automations
- Processes
- Jobs
- Apps
- Triggers
- Logs
- Monitoring
- Queues
- Assets
- Storage Buckets
- Test Suite - Orchestrator
- Resource Catalog Service
- Authentication
- Integrations
- Classic Robots
- Troubleshooting
Orchestrator Read-only Replica
Orchestrator maintains a read-only replica of its read-write operational database, with the main purpose of improving performance. Some read-only workloads are directed towards the read-only replica, so as to optimize write-replica consumption. This allows data to be read and loaded faster, which, in turn, ensures the performance of your system.
As a result of this implementation, a slight delay between the write and read actions might occur. This is to be expected, and has no impact on data becoming available.
An example of this scenario is retrieving a queue item as soon as it was added: when performed from the interface, the new queue items might not be displayed instantly, as there is a slight asynchronicity between the operational database (where the item is written) and the replica (which is called for reading the item). You can, however, avoid this by using the Get transaction item activity instead of Get queue items, since the former includes the queue item in its response.