buck.woody
LinkedIn | FaceBook | Twitter
Resume
Before we get started, let me tell you where these reports are. We are still in the server-level reports, which span all the activity on a system. You can find these reports by starting SQL Server Management Studio and right-clicking an instance name. From the menu that appears, click the “Standard Reports”, and then select one of these:
These reports are useful to identify and track down issues related to long-running queries. Normally queries create performance degradation in one of two (or both) components: the CPU and the IO system. The IO system is simply the transfer of data back and forth to the storage unit. This might mean there is pressure on the controller, the SAN, the local drive or even the path between them all. Either way, you’re looking for a spike in the graph that you can dig into for more info, all arranged by the query number shown on the graph.
As I mentioned, the reports contain the same columns but slightly different graphics, so I’ll describe them once here, and explain what each of the columns mean.
The first section contains two graphs. They show an average of the object being measured, and the other shows the total count of the object. Which one comes first is determined by the report you’re looking at.
After the graphics, the first band of the report has two columns:
Selecting the plus-sign next to the SQL text expands the data, and contains the following info:
PingBack from http://www.absolutely-people-search.info/?p=115
Buck,
The links for the two sentences:
# Logical Reads The number of logical IO reads the statement has taken so far. More about that here.
# Logical Writes The number of logical IO writes the statement has taken so far. More about that here.
Do not seem to work
Sorry about that, Randy - I'll look those up and fix them! Thanks for the catch.
CPU time ? IS it milli or micro seconds?