1. Monitoring readiness:
Monitoring party:
1) Install NetBIOS under the TCP/IP protocol
2) Login with administrator
Monitored by:
1) Monitored windows to turn on two services:
Remote Procedurecall (RPC) and remote Registry Service
Open Location: Right-click My Computer, services and applications, services
2) C $ shared folder in Windows to be monitored
Check location: Right-click "My Computer", shared folders----Share with C $ this shared folder, if you do not manually add
3) Monitored Windows has administrator privileges, including user name and password
4) Modify the network security settings:
(Group Policy gpedit.msc Windows Settings-security settings-Local Policies-security options-right network access-local account sharing and security mode-> modified to "classic", Default is "Guest only", this corresponds to registry key: "ForceGuest" =dword:0 0000000;
Note: This step can not operate first, if the LR added Windows monitoring, will error when the operation, you can form a. reg file yourself, such as Test.reg, and then double-click to modify the corresponding registry value. Test.reg.reg specific input content is:
Windows Registry Editor Version 5.00
[Hkey_local_machinesystemcurrentcontrolsetcontrollsa]
"ForceGuest" =dword:00000000
or by manual modification.
5) Verify that the connection is possible:
Use run on the machine where LR is installed. Enter the monitored machine ipc$, such as 192.168.133.99c$ and then enter the Administrator account number and password, if you can see the monitor machine's C drive, it means that you get the machine's administrator rights, you can use LR to connect, but this step may be prompted " Unable to find the network path ", you can try to add monitoring to the LR above, to see the effect, OK.
2. Monitoring Windows resources with LR
This is simple, add the measure directly, if there is a window that pops up the Windows Resource Selection object is successful:
Mouse Select Windows Resource Monitoring window, click on the right button pop-up menu and select "Addmeasurements ..."
Click "Add" to enter the IP address of the monitored server, click OK
If you can go online to the server, you will see all the counters in the resource metrics, if the point "OK" then the system is selected by default, all performance curves are displayed in the Watch window, you cannot filter the display of a curve individually, and if a counter is selected, the point "add" pops up other performance indicators under the item. Select the desired counter and then click "Add"
It is important to note that the user you log into the client (that is, you have a loadrunner machine) should be an administrator, while also ensuring that the user is an administrator on the monitored server. This option, although all performance curves are still displayed in the Watch window, can be displayed by right-clicking the popup menu and selecting a curve that you specify. The method is to double-click the Watch window to enlarge the display, and then right-select "Show only the specified graph" Monitoring window can also be superimposed and other operations, powerful, through the right-click menu selection can be complex display operation. Commonly used is the Web application server diagram, database server resource map, and so on, add a method similar.
3. Windows Resource Monitoring parameter analysis
Object |
Measure |
Describe |
System |
%total Processor Time |
The average percentage of time that all processors on the system are busy executing non-idle threads. On multiprocessor systems, this value is 100% if all processors are always busy, and if all processors are 50% busy, this value is 50%, and if one-fourth of those processors are 100% busy, this value is 25%. It reflects the ratio of time spent on useful jobs. Each processor is assigned to an idle thread in the idle process to consume those unproductive processor cycles that are not used by all other threads |
System |
File Data operations/sec |
The speed at which the computer issues read and write operations to the file system device. This action does not include file control operations |
Processor |
%processor time (Windows 2000) |
The percentage of time that the processor is executing a non-idle thread. This counter is designed as a primary indicator of processor activity. It is the amount of time it takes to measure the processor's threads executing idle processes in each sampling interval. This value is then subtracted from the total amount of time to be computed (each processor has an idle thread that consumes the remaining CPU resources when no other thread is running). It can reflect the percentage of sampling intervals that a useful job occupies. This counter shows the average percentage of busy times observed during sampling. He 2 is calculated by monitoring the time that the service is inactive, and then subtracting this value from the total amount of time |
System |
Processor Queue Length |
The immediate length of the processor queue in the thread unit. This technique is always 0 if the thread count is not monitored at the same time. All processors use a single queue (where threads wait for the processor to loop through the queue). This length does not include the thread that is currently executing. In general, if the processor queue length has been more than 2, it may indicate a processor blockage. This value is an instant count, not an average for a period of time |
Memory |
Page faults/sec |
This value is the count of page faults in the processor. A page fault occurs when a process references a specific virtual memory page that is not in the working set of its main memory. If a page is in the standby list (so it is already in main memory), or it is being used by another process that is sharing the page, the page fault does not cause the page to be fetched from disk |
PhysicalDisk |
%disk time |
Percentage of time spent on read-write requests served by selected disk drives |
Memory |
Pool nonpaged Bytes |
Number of bytes in the nonpaged pool, which is the area of the system memory from which the control is obtained when the operating system component finishes the specified task. Nonpaged pool pages may not exit into the paging file. They are always in main memory since they were allocated |
Memory |
Pages/sec |
The number of pages read from disk or pages written to disk for parsing memory references to pages (which are not in memory when referenced). This is the and of "pages Input/sec" and "pages Output/sec". The page traffic included in this counter represents the system cache that is used to access the file data for the application. This value also includes pages that are passed to/from non-cached mapped memory files. This is the main counter to consider if you are concerned about the memory pressure problem (i.e. system failure) and the excessive paging that may be generated. |
System |
Total Interrupts/sec |
The speed at which the computer accepts and handles hardware interrupts. Devices that may generate interrupts have a system clock, a mouse, a data communication line, a network interface card, and other peripheral devices. This count indicates how busy these devices are on the computer |
Objects |
Threads |
The number of threads the computer has in the cell phone data. Note that this is an instant count, not an average for a period of time. A thread is the basic executable entity used to execute instructions in the processor |
Process |
Private Bytes |
The current number of bytes that are specifically allocated for this process and cannot be shared with other processes |
LR Monitoring Windows Resources