Windows NT Performance Monitor
To create and analyze a baseline, performance monitor should be used with a database application such as Microsoft Access according to Microsoft.
Performance Monitor Objects
- Cache - Level 2 cache. Data Map Hits %
- Logical Disk* - % Free Space
- Memory* = Pages/Sec
- Objects - Process and thread counts
- Paging file - Virtual memory. %Usage
- Physical disk - Disk Queue Length, Average disk Sec/Transfer
- Process - Currently running programs.
- Processor* - %Processor Time
- Server - Bytes Total/Sec
- System - NT Performance. File Read or Write Operations/Sec
- Thread - Thread performance
Performance Monitor can monitor remote computers. Performance Monitor selections include:
- Computer - The local or remote computer to monitor.
- Object - Which object to monitor
- Counter - A counter associated with the object. There are several counters associated with each object.
Running Performance Monitor
Select "Start", "Programs", "Administrative Tools", "Performance Monitor" to run the performance monitor tool.
Monitoring Processor Performance
- Processor - %processor time - The processor is not a bottleneck until %processor time is sustained at 80% or more. Spikes to 100% are normal. Multiple processors help if running more than one process or a application uses multiple threads.
- Processor - Interrupts/second - Should be 100 to 1000 with spikes to 2000. If too high, a hardware device may be malfunctioning.
- System - Processor queue length - Should be less than 2. To monitor process queue length a thread counter must also be monitored by selecting "Edit" in the menu, then "Add to chart", then select "Thread" in the object drop down list and select a measurement parameter like "context switches/sec".
Measuring Disk Performance
To measure disk performance, disk counters must first be enabled since they are disabled by default due to their cause of performance loss. Disk counters are enabled by entering "diskperf -y" on the command line. After issuing the "Diskperf -y" command, reboot to make the change effective. If using a stripe disk set, use the command "diskperf -ye". To turn off disk counters, enter "diskperf -n" on the command line. The following characteristics show disk performance:
- Memory: Pages per second - Pages going between physical memory and the paging file. Should be less than 20.
- Memory: Available bytes - Virtual memory availability. Should be greater than 4Mb.
- Memory: Committed bytes - Should be less than physical memory.
- Memory: Page faults/sec - This indicates the number of times pages were not in RAM and needed to be called up from the hard drive.
- Logical Disk: % disk time - This indicates the amount of processor time spent serving disk requests. Measured against Processor :% processor time, it will indicate whether processor time is being used up by disk requests. If over 90% the disk or the controller is a bottleneck.
- Logical disk: Disk Bytes per transfer - Indicates the speed of the disk drives.
- Logical Disk: Current Disk queue length
- Logical disk: Disk bytes per second
Logical disks operate at a higher level than physical disks since logical disk performance mapping may include network shared drives.
Measuring Network Performance
- Network Card: % network utilization - The network monitor agent must be installed and running. This should be no greater than 30% or 40% on ethernet networks and 80% on token ring networks.
The control panel network applet allows the following network optimizations:
- Minimize memory used - For less than 10 users.
- Balance - For 10-64 users.
- Maximize Throughput for File Sharing - Default. 64 or more users. Memory is allocated for file sharing.
- Manage Throughput for Network Applications - To support distributed applications.
Ways to View Statistics
- Alerts - The administrator can be notified when a counter exceeds or falls below a preset value. Saved as *.pma file. The computer for the alert along with object, counter, and threshold value must be specified. A specified program may be run if an alert is triggered.
- Chart - The default view. Graphs and histograms (vertical bar charts) are used. To switch to histograms, use the menu item "Options", "Chart" selection. The Gallery section has Graph and Histogram radio button selections. Graphs display data every second and display 100 seconds worth of data. Chart file saved as *.pmc. A particular counter may be highlighted by clicking on the counter and pressing the backspace key. Chart Options include:
- Legend checkbox
- Value Bar checkbox
- Vertical Grid checkbox
- Horizontal Grid checkbox
- Vertical Lables checkbox
- Vertical maximum textbox (Default is 100).
- Gallary section - Graph and Histogram checkboxes
- Update Time section - Periodic Updates and Manual Updates checkboxes with Interval (seconds) textbox.
- Log - Used to create data in log files for future analysis. Data can be acquited from several systems in one log file. Log files can be used to create charts, reports, or alerts by sending them back through performance monitor. Saved as *.pml. This information may be exported to a spreadsheet or database.
- Report - Used to show a large number of objects and counters at one time. It is a list of counters and their average values. Saved as *.pmr.
- Processes - Shows PID, CPU, CPU time, and memory usage.
- Performance - Shows:
- CPU usage and history
- memory usage and history
- Total handles, threads, and processes
- Physical memory
- Commit Charge - Memory allocated to the system or programs.
- Kernel memory