You are evaluating the migration of an on-premises Microsoft SQL server database to a Microsoft Azure SQL database. You need to identity the minimum service tier required to host the database. Which…

The number of concurrent sessions or requests active on the database is a critical metric for sizing, as higher concurrency often requires more DTU/vCore resources and can impact performance limits across various service tiers.

I/O latencies are crucial because database performance heavily relies on storage I/O throughput and responsiveness. Different Azure SQL Database service tiers offer varying I/O performance characteristics, so matching required latencies ensures adequate post-migration performance.

The number of worker threads is an internal SQL Server metric related to how requests are handled; however, 'concurrent sessions' is a more direct and high-level metric for determining the overall performance tier requirements.

The maximum degree of parallelism (MAXDOP) configured and utilized by queries on-premises indicates the CPU-intensive nature of workloads and the demand for parallel processing. This metric helps determine the necessary compute capacity (vCores or DTUs) in an Azure SQL Database service tier.

Database size primarily impacts the storage cost and storage capacity limits of a service tier, but it is not the main metric for determining the *minimum performance tier* required to host the database's workload characteristics.