Update azure-sql/virtual-machines/windows/storage-performance-analysis.md

Author: JillGrant615

azure-sql/virtual-machines/windows/storage-performance-analysis.md

@@ -160,7 +160,7 @@ The following lists potential causes for latency without throttling:
 - **Background processes**: Other processes on the VM - such as antivirus software, backups, or maintenance tasks (like Windows Update) - can consume CPU, memory, or disk I/O resources, which delay SQL Server operations. Inefficient filter drivers can worsen this effect. These processes compete with SQL Server for system resources, causing I/O delays that appear as storage latency. For instance, an antivirus scan that reads numerous files simultaneously can reduce disk bandwidth available to SQL Server, which increase latency for database transactions. Additionally, not having the right antivirus exclusions can introduce latency issues without throttling in SQL Server on Azure VMs, primarily through increased disk I/O, filter driver interference, and resource competition.
 - **Lower tiered storage usage**: Opting for lower-tier storage options, such as Standard HDDs, instead of Premium SSDs or Ultra Disks, introduces higher baseline latency due to the inherent design of these disks, even without hitting IOPS limits. While cost-effective, lower-tier storage is not optimized for performance-sensitive SQL Server workloads, which leads to slower data access. For example, a customer using Standard HDDs to save costs might experience slower query performance due to the disks' naturally higher latency.
 - **Inadequate storage configuration**: Failing to configure your storage to optimize for SQL Server workloads can lead to latency without throttling. For example, incorrect disk caching settings can degrade performance. Microsoft recommends enabling read-only caching for data disks and disabling caching for log disks when using Premium SSD v1 with SQL Server on Azure VMs. Misconfigured caching can slow down read or write operations. For example, disabling read caching on a data disk that hosts SQL Server data files reduces the efficiency of read-heavy workloads, increasing latency.
-- **SQL Server database contention**: Inefficient queries (for example, a full table scan instead of an indexed lookups) or lock contention within SQL Server can increase I/O demand or delay data access, which manifests as storage latency. Application-level issues can strain the storage subsystem without exceeding its limits, particularly with small, random I/O patterns common in transactional workloads. For example, a poorly optimized query that performs a full table scan on a large dataset will read excessive data from disk, boosting I/O load and latency compared to an indexed query.
+- **SQL Server database contention**: Inefficient queries (for example, a full table scan instead of an indexed lookup) or lock contention within SQL Server can increase I/O demand or delay data access, which manifests as storage latency. Application-level issues can strain the storage subsystem without exceeding its limits, particularly with small, random I/O patterns common in transactional workloads. For example, a poorly optimized query that performs a full table scan on a large dataset will read excessive data from disk, boosting I/O load and latency compared to an indexed query.
 
 If you're experiencing latency without throttling, consider the following steps to address the latency: