The Disk panel tracks the performance of the disk used by your monitored SQL Server. Unexpected spikes in latency, throughput, or SQL Server IO may warn you about the beginning of a serious performance issue. Chronically high disk metrics can signify an insufficient IO subsystem, or excessive or inefficient IO activity on the server. Use the Disk view on the Resources tab to track several key disk statistics over a period of time. For additional information, see Get the disk performance details.
Latency chart
The Disk Latency chart plots the average time (in milliseconds) required to complete disk reads and writes, tracking the disk IO speed over time.
Metric | Why it is important |
|---|---|
Average Disk ms/Read | Higher disk read times impact the ability of SQL queries to retrieve data from your databases. |
Average Disk ms/Write | Higher disk write times impact the ability of SQL queries to update your databases, from simple text column modifications to dropping tables. |
Throughput chart
The Disk Throughput chart plots the number of disk reads and writes over time.
Metric | Why it is important |
|---|---|
Disk reads/sec | A high number of disk reads may indicate that the SQL Server is executing poorly-performing queries. |
Disk writes/sec | A high number of disk writes may indicate that the SQL Server is running low on allocated memory. |
SQL Server Physical I/O chart
The SQL Server Physical I/O chart breaks down the SQL Server throughput into its key component pieces, giving you a detailed view of how well SQL Server is processing your data requests and updates.
Metric | Why it is important |
|---|---|
Checkpoint writes | A high number of checkpoint writes may indicate that pages are not written to disk quickly enough for optimal performance. |
Lazy writer writes | A high number of lazy writer writes may indicate that the SQL Server is running low on allocated memory. |
A high number of page reads may indicate that the SQL Server is executing inefficient queries or that you should optimize the database indexes. | |
A high number of page writes may indicate that the SQL Server is executing inefficient queries or that you should optimize the database indexes. | |
Read ahead reads | A high number of read-ahead reads may indicate that you should tune your query execution plan or that you should defragment the disk. |
Available alerts
- Average Disk Milliseconds Per Read Alert
- Average Disk Milliseconds Per Write Alert
- Disk Reads (Per Second) Alert
- Disk Writes (Per Second) Alert