We are excited to announce that Azure SQL Managed Instance (MI) Business Critical now supports up to 128 vCores (virtual cores) with Premium-series hardware.
What is a vCore?
A vCore represents a unit of compute resources. It is a logical CPU that provides processing power to handle your workloads and queries.
The number of vCores in Azure SQL Managed Instance determines the amount of compute resources allocated to the instance. This directly impacts the performance and scalability of the instance; more vCores provide higher processing power, which can help in handling larger workloads and improving query performance.
128 vCores is the new limit for the number of vCores in SQL MI
In SQL MI, you can choose different hardware configurations (Standard, Premium-series, and Premium-series memory-optimized) and service tiers (General Purpose and Business Critical) that offer varying amounts of vCores.
For SQL MI Business Critical instances with Premium-series hardware configurations, we increased the maximum number of vCores to 128 vCore. The new set of vCores available to SQL Managed Instances is shown in Table 1. This allows you to select the appropriate compute resources based on your workload requirements, ensuring optimal performance and cost-efficiency.
|
Feature |
General Purpose |
Business Critical (new in blue) |
|
Number of vCores* |
4, 8, 16, 24, 32, 40, 64, 80 |
Standard-series (Gen5): 4, 8, 16, 24, 32, 40, 64, 80 |
Table 1 - Number of vCores in Azure SQL MI Business Critical tier
When can more vCores speed up your query workload?
In general, a SQL MI with more vCores can manage more concurrent connections, transactions, and queries at the same time i.e., CPU-intensive workloads. CPU-intensive SQL workloads are database operations that require a significant amount of processing power from the CPU to complete. These workloads typically involve complex queries that require extensive calculations, data aggregation, sorting, or filtering.
Examples of SQL workloads that can be CPU-intensive are:
- Aggregation queries: Queries that aggregate data from multiple tables or partitions, such as SUM, AVG, MAX, and MIN.
- Joins: Queries that join large tables or multiple tables with complex conditions.
- Sorting and filtering: Queries that require sorting or filtering large datasets, especially if the data is not indexed properly.
- Data transformation: Queries that transform or convert data from one format to another.
- Analytics and reporting: Queries that generate complex reports or analytics, such as trend analysis or forecasting.
- Indexing: Building or rebuilding indexes on large tables.
In general, any SQL workload that requires extensive processing or manipulation of large amount of data can be CPU-intensive and thereby sped up with additional vCores.
What about additional memory for additional vCores?
In SQL MI, you got used to vCore-based purchasing model (What is Azure SQL Managed Instance? - Azure SQL Managed Instance | Microsoft Learn) and fixed memory-to-vCore ratio depending on the hardware configuration.
However, for these additional vCores (values in blue), you will not get additional memory (values in red). See Table 2 and its graphical representation on Graph 1.
|
Feature |
Business Critical |
|
Number of vCores |
Standard-series (Gen5): 4, 8, 16, 24, 32, 40, 64, 80 |
|
Max memory |
Standard-series (Gen5): 20.4 GB - 408 GB (5.1 GB/vCore) on each replica Premium-series: 28 GB - 560 GB (7 GB/vCore, up to 80 vCores), fixed at 560 GB for 96 and 128 vCores, on each replica Memory optimized premium-series: 54.4 GB - 870.4 GB (13.6 GB/vCore, up to 64 vCores), fixed at 870.4 GB for 80, 96 and 128 vCores, on each replica |
Table 2 - Number of vCores and maximum memory in Azure SQL MI Business Critical tier
Graph 1 - SQL MI Business Critical: Max Memory for different number of vCores
Conclusion
This improvement in Azure SQL Managed Instance allows you to migrate more of your CPU-intensive SQL workloads to Azure SQL MI. If you're still new to Azure SQL Managed Instance, now is a great time to get started and take Azure SQL Managed Instance for a spin!