LoadRunner processor, memory, network interface performance object common performance counters

TPS 1 transactions per second is also the number of tasks/second

2. Abbreviation of throughtput per second. Unit: Byte/second
Bytes/second, that is, throughput .....

[Share] network interface counter

 

Many people misunderstand the speed units such as kbps, kb, and Mbps,
The following briefly explains how to calculate the so-called 1.5 m, 3 M, 6 m.
The so-called 1.5 M bandwidth actually refers to 1.5 Mbps (bits per second), that is, 1.5x1024/8 = 192kb/sec,

However, this is only the theoretical speed. In fact, we need to deduct about 12% Ethernet header, IP header, TCP header, and ATM.
Header and other control signals, so the transmission speed limit should be around 169kb/sec.

In terms of transmission unit writing, B and B represent bytes and bits, respectively. The definitions of the two are different.

1 byte = 8 bits
1 kb = 1024 bits
1 kb = 1024 bytes
1 MB = 1024 KB
1 MB = 1024 KB

Maximum bandwidth download theoretical value

1.5 m = 169 kb/s
3 M = 338 KB/s
6 m = 676 kb/s
10 m = 1126 kb/s

The above is a theoretical value. You can test the actual connection speed by downloading files,

Bytes total/sec is the speed at which bytes are sent and received on each network adapter, including frame characters. Network
Interface // bytes total/sec = Network Interface // bytes encoded ED/sec + network
Interface // bytes sent/sec.

Current bandwidth refers to the current bandwidth of the network interface estimated by bit/second.

The output queue length is the length of the output data queues (data packets. If the length is longer
2. it will delay and, if possible, identify and solve the bottleneck problem. Because the request is queued by the Network Driver Interface Specification (NDIS) in this operation, this will always be 0.

Packets/sec is the rate at which packets are sent and received on the network interface.

Packets outbound discarded
Select the number of dropped output packets, even if no error is found that will block the transmission of these packets. The possible reason for discarding data packets is to release the buffer space.

Packets outbound error indicates the number of output data packets that cannot be transmitted due to an error.

Packets inclued discarded
Indicates the number of input packets to be discarded, even if no error is found that hinders these packets from being passed to higher-level protocols. The possible cause for discarding data packets is to release the buffer space.

Packets received error indicates the number of input packets, which contain errors that prevent them from being passed to higher-level protocols.

Packets received ED/sec is the speed at which packets are received on the network interface.

Packets sent/sec is the rate at which data packets are sent on the network interface.

 

 

[Share] processor counter
Processor counters

% Processor time
The percentage of time that the processor uses to execute non-Idle threads. The calculation method is to measure the non-idle thread activity time within the sample interval, and subtract this value from the sample interval. (Each processor has an idle thread in
There are no other threads that can run the time consumption cycle ). This counter is the main indicator of processor activity and displays the average percentage of busy time observed at sample intervals. This value is 100%
Calculated based on the time when the service is not active. Generally, the average CPU activity should be below 80%. exceeding 80% indicates that the CPU processing capability has reached the limit.

% DPC time refers to the percentage of time that the processor receives and provides services during the sample interval that is used by the deferred program call (DPC. DPC
The running interval is lower than the standard interval priority. Because DPC is executed in privileged mode, DPC
The percentage of time is part of the percentage of privileged time. These times are calculated separately and are not part of the total number of interval computations. This total number shows the average busy hours as the percentage of instance time. The lower the value, the better. In Multiple
In the processor system, if the value is greater than 50% and processor: % Processor
Time is very high. Adding a network adapter may improve performance and the network provided is not saturated.

% Privileged time the percentage of time it takes to process the code executed by the thread in privileged mode. When Windows
During System Service, this service often runs in privileged mode to obtain access to system proprietary data. Threads executed in user mode cannot access the data.
Calls to the system can be direct (explicit) or indirect (implicit), such as page errors or interruptions. Unlike some early operating systems, Windows
In addition to the traditional user and privileged mode, the processing boundary is also used as the subsystem protection. Some
Operations performed for your application may not only appear in the processing privileged time, but also appear in other subsystems. This time includes the CPU maintenance interruption and delay process call time. If the value is too high
I/O processing causes a large number of system interruptions.

% USER time
The percentage of time when the processor is in user mode. User mode is a finite processing mode designed for applications, Environment subsystems, and integer subsystems. Another mode is the privileged mode, which is designed for operating system components.
And allow direct access to hardware and all memory. The operating system converts the application thread to the privileged mode to access the operating system service. This Count value is displayed as part of the sample time during average busy hours.

Interrupts/sec
The average speed at which the processor receives and processes hardware interruptions. The unit is the number of instances per second. This does not include the latency of separate counting process calls (dpcs ). This value indicates the device that is interrupted (such as the system clock, mouse
Activities of labels, disk drives, data communication lines, network interface cards, and other peripheral devices. These devices usually interrupt the processor when they complete the task or need attention. Normal thread execution is interrupted. System clock
Interrupt the processor every 10 milliseconds to create the background of the interrupt activity. This Count value shows the difference between the values observed in the above two instances except the value obtained from the duration of the Instance interval.

% Interrupt time is the time when the processor receives and the service hardware is interrupted during the instance interval. This value indirectly indicates the device activity that generates the interval,
Such as system clock, mouse, disk driver, data communication line, network interface, and other peripheral devices. When these devices complete a task or need to be managed, they usually interrupt the processor. Normal during interruption
The thread execution will stop. Most system clocks run at intervals of 10
Interrupt the processor in milliseconds to generate the background of the interval activity. During the interval, the normal thread execution is terminated. This counter shows that the average usage time is part of the Instance time.

Private bytes indicates the number of allocated bytes that cannot be shared with other processing.

Page faults/sec
The speed at which page errors occur due to execution of threads in this process. When the thread references a virtual memory page that is not in the master memory working set, the page appears.
Fault. If it is being used in a backup table (that is, it is already in the primary memory) or another shared page, it will not be able to retrieve pages from the disk.

% USER time
The percentage of time that the processing thread uses to execute code in user mode. Applications, Environment subsystems, and collection subsystems are executed in user mode. Windows
The executable program, kernel, and device driver of is not damaged by code executed in user mode. Unlike some early operating systems, Windows
In addition to the traditional user and privileged mode, the processing boundary is also used as the subsystem protection. Some
Operations performed for your application may not only appear in the processing privileged time, but also appear in other subsystems.

% Privileged time is the percentage of time taken to process the code executed by the thread in privileged mode. When Windows
During System Service, this service often runs in privileged mode to obtain access to system proprietary data. Threads executed in user mode cannot access the data. System calls can be direct.
(Explicit) or indirect (implicit), such as page errors or intervals. Unlike some early operating systems, Windows
In addition to the traditional user and privileged mode, process boundaries are also used as subsystem protection. Some
Operations performed for your application may not only appear in the privileged time of the process, but also in other subsystem processes.

% Processor time
The percentage of time taken by all process threads to execute commands using the processor. Commands are the basic unit for computer execution. A thread is the object for executing commands, and a process is the object created when the program is running. This count includes processing
Code executed by hardware intervals and trap conditions.

Virtual bytes
The current size of the virtual address space used for processing in bytes. The use of virtual address space does not necessarily refer to the use of disks or primary memory pages. Virtual space is limited, and the number of loads may be limited.
Database capabilities.

Working Set refers to the number of current bytes in the processing working set. Working Set
Is the Memory Page set recently touched by the thread during processing. If the available memory on the computer is above the threshold, even if the page is not in use, it will remain in a working
Set. When the available memory falls below the threshold, the page will be deleted from the working set. If you need a page, it will return a soft fault to the working set before leaving the main memory.
.

Page file bytes indicates the maximum number of bytes used for processing in paging file. Paging file
Used to store memory pages that are not included in other files for processing. Paging file is shared by all processes, and paging file
Insufficient space will prevent other processing from allocating memory.

I/O data bytes/sec processing from I/O
Speed of reading/writing bytes. This counter counts all activities generated by this processing, including files, networks, and device I/O.

 

 

[Share] memory counters

Page faults/sec
Average number of error pages per second. Because there is only one page error in each error operation, the calculation unit is the number of error pages per second, which is equal to the number of page error operations. This counter contains hard errors.
(Those requiring disk access) and soft errors (error pages found elsewhere in the physical memory ). Many processors can continue to operate in the case of a large number of soft errors. However, hard errors can lead to significant delays.
Extension. A page error occurs when a process requests a piece of memory but the system cannot allocate it. This value is too high (compared with that without pressurization) for two reasons:
1. The application occupies too much memory, which can be solved by increasing the amount of memory.
2. The memory requests of applications are too frequent (for example, frequent object creation and destruction ). Consider changing the design.

Committed bytes
The confirmed virtual memory in bytes. Check that the memory disk page file retains the physical memory of the space. Each physical disk can have one or more page files. This counter only displays the last observation.
It is not an average value.

The available physical memory size of processes running on the available Mbytes computer, measured in kilobytes instead
Memory // available bytes. It is calculated by adding zero, idle, and the space in the backup memory list. Idle memory can be used at any time;
Zero memory is used to prevent future processes from seeing the data used by the previous process and fills up zero memory in many pages of memory. The standby memory refers to the working set of the slave process (its physical
Memory) is moved to the disk, but the memory can still be called again.
This counter only displays the last observed value; it is not an average value. When the value is changed to an hour, Windows frequently calls disk page files. If the value is small, for example, less than 5
MB, the system will spend most of the time on the Operation page file.

Pages/sec
The speed at which hard page errors are read from or written to the disk. This counter is the main indicator that can display the type error that causes system range delay. It is memory // pages
The sum of input/sec and memory/pages output/sec. The number of pages is calculated so that the page count can be the same as that of other pages without conversion:
Memory // page faults/sec
For comparison, this value includes pages retrieved from non-Cache ing memory files cached by the File System (usually requested by applications) to meet errors.
Generally, if pages/sec continues to exceed several hundred, you should study the page exchange activity further.

Commit limit
The amount of virtual memory that can be used without the need to expand paging files. This is calculated in bytes. The confirmed memory is the physical memory retained on the disk paging file. Each Logical Disk can have one
Paging memory. If you expand the paging file, this limit will be increased accordingly. This counter only displays the value observed in the previous review, rather than an average value.