A new Azure Analysis Services setting (in preview) improves performance and consistency of query-replica synchronization, delivering more reliable analytics experiences at scale.
Background
Azure Analysis Services is a fully managed platform as a service (PaaS) that provides enterprise-grade data models in the cloud. When you scale out an Analysis Services server, client queries can be distributed among multiple query replicas in a query pool, reducing response times and improving availability during heavy query workloads.
The Problem with Previous Synchronization
Previously, when a processing operation completed, there could be a brief window where some query replicas had the updated model while others still served queries against the older version. This inconsistency could lead to different users receiving different results for the same query during the synchronization window.
The New Synchronization Improvement
The new preview feature introduces a parallel sync mode that synchronizes all query replicas simultaneously after a processing operation completes. Key improvements include:
- Consistency: All replicas switch to the new model version at the same time, eliminating result inconsistencies.
- Performance: Synchronization time is reduced because replicas sync in parallel rather than sequentially.
- Monitoring: New metrics in Azure Monitor allow you to track synchronization status and duration.
How to Enable the Preview
The new synchronization behavior can be enabled through the Azure portal, PowerShell, or the REST API by setting the queryreplica.SyncMode property to Parallel.
Implications for Enterprise Analytics
For organisations running mission-critical BI workloads on Azure Analysis Services, this improvement reduces the risk of stale data being served to end users after data refreshes, improving trust in the analytics platform.