LoadRunner Performance Test Results Counter indicator description

LoadRunner Performance Test Results Counter indicator descriptionreprinted 2015-09-23 09:57:13 Tags: loadrunner counter

For the performance test results analysis process, the face of a large number of test data, but feel not to analyze. Today we have a simple description of the individual monitored objects in the Windows operating system counters.

Memory:

· Available Mbytes

Description: The number of physical memory available. If the value of available MBytes is small (4 MB or less), the total memory on the computer may be low, or a program does not release memory.

Reference value: 4 MB or less, with at least 10% physical memory value

-----------------------------------------------------------------------------------------------?

· Page/sec (Input/out)

Summary: Number of pages fetched or written from disk in order to parse hard page faults.

Generally if the page/sec continues above hundreds of, then you should study the paging activity further. It may be necessary to increase the memory to reduce the need for paging (you can multiply this number by 4 K to get the resulting drive data traffic). A large value of pages/sec does not necessarily indicate a memory problem, but may be caused by a program running a memory-mapped file.

-----------------------------------------------------------------------------------------------?

· Page Fault

Description: Error pages processed per second by the processor (including soft/hard errors). When the processor requests a page (possibly data or code) for an error in the location specified by the memory, this forms a page Fault. If the page is elsewhere in memory, the error is referred to as a soft error (measured with the transition fault/sec register), or a hard error if the page must be re-read from the hard disk. Many processors can continue to operate in the event of a large number of soft errors. However, a hard error can lead to noticeable delays.

-----------------------------------------------------------------------------------------------?

· Page input/sec

Description: The number of pages read from disk in order to resolve the hard error page.

-----------------------------------------------------------------------------------------------?

· Page reads/sec

Summary: Number of reads from disk in order to resolve hard error pages. Resolves a reference to memory, the number of times the page file must be read. The threshold value is >5. The lower the better. Large values indicate disk reads instead of cached reads.

-----------------------------------------------------------------------------------------------

· Cache Bytes

Description: The file system cache, which by default is 50% of the available physical memory. If IIS5.0 runs out of memory, it automatically organizes the cache. You need to focus on the trend change of the counter. The indicator shows only the last observed value, which is not an average.

-----------------------------------------------------------------------------------------------

Pool Paged Bytes

Summary: Refers to the number of bytes in a paged pool, which is an area of system memory that can be used by an object.

-----------------------------------------------------------------------------------------------

Pool Nonpaged Bytes

Summary: Refers to the number of bytes in the nonpaged pool, the number of bytes in the nonpaged pool if the persistent increase indicates a possible memory leak problem, additional metrics need to be combined to determine whether there is a serious memory leak or other cause of the nonpaged pool increase.

-----------------------------------------------------------------------------------------------

• Memory leaks

Summary: If you suspect a memory leak, monitor the memory\\ Available Bytes and memory\\ Committed Bytes to observe the memory behavior and monitor the process\\private Bytes of processes you think might be leaking memory, Process\\working set and Process\\handle Count.

If you suspect that a kernel-mode process is causing the leak, you should also monitor Memory\\pool nonpaged Bytes, memory\\ Pool Nonpaged Allocs, and process (process_name) \ Pool Nonpaged Bytes.

===================================================

Process

· Page faults/sec

Description: Compare the page faults generated by the process with the system generation to determine the impact of this process on the system page failure. The average number of page errors per second.

-----------------------------------------------------------------------------------------------

· Private Bytes

Description: The current number of bytes allocated by this process that cannot be shared with other processes. If system performance decreases over time, this counter can be the best indicator of memory leaks.

-----------------------------------------------------------------------------------------------

· Work set

Description: Handles the most recent memory pages used by threads, reflecting the number of memory pages used by each process. If the server has enough free memory, the page is left in the working set, and when free memory is less than a specific threshold, the page is purged from the working sets.

===================================================

Processor

% Processor Time

Summary: CPU utilization, you can see whether the processor is saturated, the best range of this value is 75%-95%, if in the performance monitoring process this value is too low to indicate that the CPU is not fully utilized, but also requires a greater load pressure, if the value continues to exceed 95%, the current system bottleneck is the CPU, Consider adding a processor or a better-performing processor at this point.

Reference value: <80%?

-----------------------------------------------------------------------------------------------

· Processorqueue Length

Summary: Determine the CPU bottleneck, if the queue length shown by Processor queue lengths remains constant (>=2) and the processor utilization%processor time exceeds 90%, there is likely a processor bottleneck. If you find Processor Queue Length shows more than 2 queues, and processor utilization has been low, perhaps more to address the processor blocking problem, where the processor is generally not a bottleneck.

-----------------------------------------------------------------------------------------------interrupt/sec

Summary: Refers to the processor receives and maintains the average of hardware interrupts, in units of events/sec, which describes the activity of devices that can produce interrupts, such as system clocks, mice, disk drives, network cards, and other external devices, which are often disruptive devices that interrupt the processor when a task is completed.

-----------------------------------------------------------------------------------------------

%user Time

Description: Refers to the percentage of time that the processor is in user mode, which is the CPU time consumed by the non-kernel operation user process. The value can be expressed as the time spent by the CPU's database operations (such as lookups, sorting, and so on), and if the value is high, consider increasing the index, using a simple table join, and splitting the large table horizontally to reduce the value.

The computer processor has the user mode and the privilege mode two modes of operation, the user mode is the limited processing mode which designs for the application, the environment sub-system and the integer sub-system; The privileged mode is designed for operating system components, allowing direct access to hardware and all memory. The operating system translates the application thread into privileged mode to access the operating system server.

-----------------------------------------------------------------------------------------------

%privileged Time

Summary: Refers to the percentage of time spent processing threads executing code. If the parameter value and the "Physical disk" value have been high, indicating a problem with I/O, consider adopting a faster hard drive system.

-----------------------------------------------------------------------------------------------

%interrupte Time

Description: The time at which the processor received and serviced hardware interrupts during the instance interval. This value indirectly represents the activity of the device that caused the interruption, such as the system clock, mouse, disk drive, network card, and other external devices, which typically interrupt the processor.

-----------------------------------------------------------------------------------------------

%d PC time

Description: The percentage of time that the processor used to receive and service a delay program call (DPC) during the interval of the instance, that is, the time spent on the network processing, the lower the better. Because DPC is performed in privileged mode, the percentage of DPC time is part of the percentage of privileged time.

-----------------------------------------------------------------------------------------------

· Queue Length

Summary: A counter that tracks the current length of the server's work queue, which shows the processor bottleneck. A continuous queue length of more than 2 indicates processor congestion may occur.

===================================================

Physical Disk

%d Isktime

Summary: Refers to the percentage of time that the selected disk drive is busy servicing a read or write request.

Normal <10, this value is too large to take too much time to access the disk, consider increasing memory, replacing faster drives, and optimizing the algorithms for reading and writing data. If the value continues to exceed 80 (the processor and network connections are not saturated at this time), a memory leak can occur.

-----------------------------------------------------------------------------------------------

· Currentdiskqueuelength

Description: The average number of read and write requests (queued for the selected disk in the instance interval). (Number of disks 1.5-twice times)

-----------------------------------------------------------------------------------------------

· Avg.Disk Queue

Length

Avg.Disk Read

Queuelength

Avg.Disk Write

Queuelength

Disk read/sec

Disk write/sec

Description: The average number of read and write requests (queued for the selected disk in the instance interval).

Disk Bottleneck judgment formula:

Number of I/Os per disk = (number of reads + (number of days))/number of disks.

If the calculated number of I/O per disk is greater than the disk's processing power, there is a bottleneck in the disk.

Reference value:

Avg.diskqueue length Normal <0.5, this value is too large to indicate that the disk IO is too slow to replace the faster hard disk.

===================================================

Comprehensive judgment:

1, Processor Queue congestion judgment method: If the processor queue length is greater than 2, and the processor utilization has been very low, there is a processor jam.

2, Processor bottleneck method: After the memory factor is excluded, if%processor time continues to be greater than 90%, and the value of%interrupt duration is greater than 15%, and the value of the NIC and hard disk is lower, it can be concluded that the processor is overloaded, unable to meet the business growth needs, The processor is the system bottleneck point.

3. Monitor out-of-memory conditions can be monitored by page/sec,available Mbytes, page read/sec, page faults/sec, and other counters, and can be measured by using the frequency of "page swapping".

Page swapping is the process of moving fixed-size code and blocks of data from RAM to disk using units called pages, freeing up temporarily unused space, which is the empty surface of the hard disk used by the operating system for virtual memory. Operating system for virtual memory mainly set two points, that is, the size of the memory paging file and the location of the paging file, the size of the memory paging file is to set the minimum and maximum amount of virtual memory, and the page location is to set virtual memory to use which partition of hard disk space.

Frequent page swapping reduces system performance, which indicates that there is not enough memory if the system "page swapping" is frequent. Reducing page swapping with tuning configuration will significantly improve system responsiveness.

4. To determine if there is a memory problem with the PAGES/SEC indicator, if the pages/sec continues above hundreds of, it may be necessary to increase the memory to reduce the need for paging, and further study the paging activity. If the pages/sec indicator is too high (hundreds of), and the hard drive data flow is not high (hundreds of kb/s) can be determined to be out of memory problems, if the pages/sec indicator is higher (hundreds of), and when the hard disk data traffic is also very high (thousands of kb/s), it can be determined that a disk problem.

5. By available MBytes to determine if there is a serious memory leak problem, if the value is small (<4m), the total memory on the computer may be insufficient, or a program always consumes without releasing memory, the system has a serious memory leak problem.

6. If the page read Operations rate page Reads/sec metric is low and the value of%disk time and Avg.Disk Queue Length is high, it is determined to be a disk bottleneck, but if the avg.sidk queue length increases while the page Reads/sec the page read rate metric is not lowered, it is determined to be out of memory.

Note: You can open the native monitor by entering Perfmon in run

Reprint: http://blog.sina.com.cn/s/blog_a450e8e50102x8dd.html

LoadRunner Performance Test Results Counter indicator description