The method and device of monitoring system performance
Technical field
The present invention relates to computer realm, in particular to a kind of method and device of monitoring system performance.
Background technology
In the course of server use, it may appear that because system resource exhaustion causes system response slow or existing without response
As.For this problem, Windows provides " performance monitoring " instrument.It can be gathered in real time, the application program in analysis system,
The performance data of service, driver etc., carrys out the bottleneck of analysis system with this, monitors the performance of component, final to help user to enter
The rational allocation of row system resource.
Current Windows performance counters are provided between hundreds of performance counters, each counter for various objects
There is no dominant contact.It is a problem that ability location-server performance bottleneck how is combined from hundreds of counters.
For being not in contact between each performance counter in the prior art, it is caused can not be according to each independent performance count
The problem of numerical value monitoring system performance of device, effective solution is not yet proposed at present.
The content of the invention
It is a primary object of the present invention to provide a kind of method and device of monitoring system performance, to solve in the prior art
Be not in contact between each performance counter, it is caused can not be according to the numerical value monitoring system performance of each independent performance counter
Problem.
To achieve these goals, there is provided a kind of monitoring system performance for one side according to embodiments of the present invention
Method.This method includes:The data collector of each part in computer system is created, each data collector preserves a prison
List is controlled, watch-list have recorded the counter identification of the performance counter associated with data collector;Each performance meter is set
Monitoring threshold corresponding to number device, constitutes monitoring threshold set;The data collector of any one or more parts is run, is passed through
Data collector reads the count value of the performance counter associated with data collector;By the counting of any one performance counter
Value and corresponding monitoring threshold are compared, and determine the performance number of each part in computer system.
To achieve these goals, there is provided a kind of monitoring system performance for another aspect according to embodiments of the present invention
Device, the device includes:Creation module, the data collector for creating each part in computer system, each data are received
Storage preserves a watch-list, and watch-list have recorded the counter identification of the performance counter associated with data collector;
Setup module, for setting the monitoring threshold corresponding to each performance counter, constitutes monitoring threshold set;Read module, is used
In the data collector for running any one or more parts, the performance associated with data collector is read by data collector
The count value of counter;Determining module, for the count value of any one performance counter and corresponding monitoring threshold to be carried out
Compare, determine the performance number of each part in computer system.
According to inventive embodiments, by creating the data collector of each part in computer system, each Data Collection
Device preserves a watch-list, and watch-list have recorded the counter identification of the performance counter associated with data collector;If
The monitoring threshold corresponding to each performance counter is put, monitoring threshold set is constituted;Run the number of any one or more parts
According to collector, the count value of the performance counter associated with data collector is read by data collector;By any one individual character
The count value and corresponding monitoring threshold of energy counter are compared, and determine the performance number of each part in computer system,
Solve and be not in contact between prior art each performance counter, it is caused can not be according to the numerical value of each independent performance counter
The problem of monitoring system performance.Realize by being combined to the counter in Windows performance counters, monitoring system performance is simultaneously
The effect of the performance bottleneck of all parts of fast location computer system.
Brief description of the drawings
The accompanying drawing for constituting the part of the application is used for providing a further understanding of the present invention, schematic reality of the invention
Apply example and its illustrate to be used to explain the present invention, do not constitute inappropriate limitation of the present invention.In the accompanying drawings:
Fig. 1 is the flow chart of the method for according to embodiments of the present invention one monitoring system performance;
Fig. 2 is the flow chart of the method for according to embodiments of the present invention one optional monitoring system performance;
Fig. 3 is the structural representation of the device of according to embodiments of the present invention two monitoring system performance;And
Fig. 4 is the structural representation of the device of according to embodiments of the present invention two optional monitoring system performances.
Embodiment
It should be noted that in the case where not conflicting, the feature in embodiment and embodiment in the application can phase
Mutually combination.Describe the present invention in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
In order that those skilled in the art more fully understand the present invention program, below in conjunction with the embodiment of the present invention
Accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only
The embodiment of a part of the invention, rather than whole embodiments.Based on the embodiment in the present invention, ordinary skill people
The every other embodiment that member is obtained under the premise of creative work is not made, should all belong to the model that the present invention is protected
Enclose.
It should be noted that term " first " in description and claims of this specification and above-mentioned accompanying drawing, "
Two " etc. be for distinguishing similar object, without for describing specific order or precedence.It should be appreciated that so using
Data can exchange in the appropriate case, so as to embodiments of the invention described herein.In addition, term " comprising " and " tool
Have " and their any deformation, it is intended that covering is non-exclusive to be included, for example, containing series of steps or unit
Process, method, system, product or equipment are not necessarily limited to those steps clearly listed or unit, but may include without clear
It is listing to Chu or for the intrinsic other steps of these processes, method, product or equipment or unit.
Embodiment 1
The embodiments of the invention provide a kind of method of monitoring system performance.
Fig. 1 is the flow chart of the method for monitoring system performance according to embodiments of the present invention.As shown in figure 1, this method bag
Include step as follows:
Step S11, creates the data collector of each part in computer system, and each data collector preserves a prison
List is controlled, watch-list have recorded the counter identification of the performance counter associated with data collector.
Specifically, being each building block in computer system by above-mentioned steps S11, a Data Collection is created
Device.The counter identification of performance counter corresponding with the building block in computer system is added in the data collector
Watch-list in.Data collector is obtained and the corresponding performance of watch-list Counter mark by calling watch-list
The count value of counter.Among practical application, the counting in watch-list can be changed according to business demand or actual conditions
Device identification entry.It is of course also possible to which by other means realize the function that watch-list is realized.
Step S13, sets the monitoring threshold corresponding to each performance counter, constitutes monitoring threshold set.
Specifically, being the performance in list according to the watch-list in each data collector by above-mentioned steps S13
Counter sets corresponding monitoring threshold, constitutes the set of monitoring threshold.Wherein, monitoring threshold is used to be directed to each performance
Counter judges the performance number of part corresponding with the performance counter.
Step S15, runs the data collector of any one or more parts, is read and is received with data by data collector
The count value of the performance counter of storage association.
Specifically, by above-mentioned steps S15, being needed according to business or actual conditions, the number of one or more parts is run
According to collector.Using data collector corresponding with all parts, the performance meter associated with each data collector is obtained respectively
The count value of number device.
Step S17, the count value of any one performance counter and corresponding monitoring threshold are compared, it is determined that calculating
The performance number of each part in machine system.
Specifically, in above-mentioned steps S17, the monitoring threshold corresponding with each performance counter set by step S13
The count value for each performance counter that value and step S15 are read, to the count value of any one performance counter and its is right
The monitoring threshold answered is compared, and determines the performance number of each part in computer system.
Above-mentioned steps S11 to step S17, by the selecting of performance counter, combine, according to each of computer system
Part or the demand according to practical business, create data collector.Performance counter is obtained by data collector
Count value, according to monitoring threshold corresponding with each performance counter is pre-set, fast and accurately judges computer system
In each part performance number.
Among practical application, the present invention combines the contact between each module of Windows operating system bottom, is counted
The combination of device can be with Fast Discovery System bottleneck.By selecting, combinatorial enumeration device, and it is monitored and analyzed, it is quick fixed
Bit processor, internal memory, the performance bottleneck of disk.
Wherein, by running operation " performance monitor " in Windows management tools, after unlatching in " data collector "
Set up processor, internal memory, the data collector of disk.In processor, internal memory has been separately added into three data collectors of disk
Targetedly different counter.
Data after data are collected, are analyzed by operational data collection device, find out systematic function bottleneck point.
Preferably, in the above embodiments of the present application, the data collector of each part includes following appoint in computer system
Meaning is one or more:Processor data collector, internal storage data collector, memory data collector.
Specifically, can be each part one data collector of establishment in computer, can also be according to business needs
Or actual conditions, it is that several parts in computer create a data collector.
Preferably, in the above embodiments of the present application, the Data Collection of each part in step S11 creates computer system
Device, each data collector preserves a watch-list, and watch-list have recorded the performance counter associated with data collector
Counter identification in, the count value of the following any one or more performance counters of the data collector of processor collection:Place
Manage device average utilization (Processor %Total Processor Time), processor using the time (Processor %
User Time), the processor break period (Processor %Interrupt Time), processor queue length (System
Processor Queue Length), wherein, the count value of any one performance counter and corresponding monitoring threshold are carried out
Compare, the step of determining the performance number of processor in computer system includes:
Step S1711, processor average utilization is compared with the first monitoring threshold, obtains the first monitoring result;
Step S1712, processor is compared using the time with the second monitoring threshold, obtains the second monitoring result;
Step S1713, the processor break period is compared with the 3rd monitoring threshold, obtains the 3rd monitoring result;
Step S1714, processor queue length is compared with the 4th monitoring threshold, obtains the 4th monitoring result;
Wherein, when any one in the first monitoring result, the second monitoring result, the 3rd monitoring result, the 4th monitoring result
When individual or multiple monitoring results are that count value is more than or equal to corresponding monitoring threshold, the performance reduction of processor is determined.
Specifically, by above-mentioned steps S1711 to step S1714, by the data collector of processor, respectively collection with
The numerical value of the performance counter of data collector association, by the comparison with monitoring threshold, determines the performance reduction of processor.
Wherein, it is probably because processor is in itself without enough abilities, or due to application program inefficiency that bottleneck, which occurs, in processor
Cause processor scarce capacity.Processor bottleneck can be confirmed using the combination of following performance counter in monitoring:
Processor average utilization (Processor %Total Processor Time):The value measurement processor is used for
Perform non-NULL idle thread has been processor average utilization with percentage of time.If this value is more than 85%, show processing
Device performance is not enough, it may be necessary to be equipped with the faster processor of speed for server.
Processor utilizes time (Processor %User Time):The value measurement processor is spent in the user mode
Used percentage of time.If this value is higher, then it represents that server is busy in handling application program.Task manager can be opened
Check that application program takes cpu busy percentage.Close or optimization takes application program of the processor compared with large resource.
The processor break period (Processor %Interrupt Time):The value was measured in the specific phase in sampling interval
Between processor be used to receiving and handling time of hardware interrupts.If the value of this counter is more than 15%, then it represents that there may be
Hardware problem.
Processor queue length (System Processor Queue Length):It represents the line in processor queue
Number of passes.If this value is often than more than twice of CPU quantity, then it represents that the processor scarce capacity of server is, it is necessary to change many
Core processor.
Preferably, in the above embodiments of the present application, the Data Collection of each part in step S11 creates computer system
Device, each data collector preserves a watch-list, and watch-list have recorded the performance counter associated with data collector
Counter identification in, the count value of the following any one or more performance counters of the data collector of internal memory collection:Virtually
Memory usage (Memory %Committed Bytes in Use), physical memory can use ratio (Memory %
Available Mbytes), nonpageable buffer pool size (Memory Pool Non-Paged Bytes), write-in disk rate
Quantity (Memory Page Read/sec) is read in (Memory Pages per Second), paging, wherein, by any one
The count value of performance counter and corresponding monitoring threshold are compared, the step of determining the performance number of internal memory in computer system
Including:
Step S1721, virtual memory utilization rate is compared with the 5th monitoring threshold, obtains the 5th monitoring result;
Step S1722, physical memory can be compared with ratio with the 6th monitoring threshold, the 6th monitoring result is obtained;
Step S1723, nonpageable buffer pool size is compared with the 7th monitoring threshold, obtains the 7th monitoring result;
Step S1724, write-in disk rate is compared with the 8th monitoring threshold, the 8th monitoring result is obtained;
Step S1725, paging reading quantity is compared with the 9th monitoring threshold, the 9th monitoring result is obtained;
Wherein, when the 5th monitoring result, the 6th monitoring result, the 7th monitoring result, the 8th monitoring result, the 9th monitoring knot
When any one or more monitoring results in fruit are more than or equal to corresponding monitoring threshold for count value, the performance drop of internal memory is determined
It is low.
Specifically, by above-mentioned steps S1721 to step S1725, by the data collector of internal memory, collection is with being somebody's turn to do respectively
The numerical value of the performance counter of data collector association, by the comparison with monitoring threshold, determines the performance reduction of internal memory.Its
In, low memory is often as RAM deficiencies, memory overflow.Normal conditions internal memory is more, and magnetic disc i/o activity is fewer, therefore
Monitoring can use following counter to diagnose possible memory bottleneck.The combination of following performance counter can be used in monitoring
To confirm memory bottleneck:
Virtual memory utilization rate (Memory %Committed Bytes in Use):It measures Committed Bytes
With Commit Limit ratio, in other words, what it was measured is the virtual memory quantity in use.If this numerical value is more than
80%, then it represents that low memory.To this, most direct solution is the more internal memories of increase.
Physical memory can use ratio (Memory %Available Mbytes):Memory %Available Mbytes
It measures the physical memory amount (unit for Mbytes) available for operation process.If this value is less than the 5% of total physics RAM,
Mean low memory, paging activity may increase.It is the result that calculates recently and not average value to note this counter.
This problem is solved, more internal memories should be increased.
Nonpageable buffer pool size (Memory Pool Non-Paged Bytes):It measures the big of nonpageable buffer pool
Small (unit is byte).But as long as what is stored in this Installed System Memory region is that can not write disk and distribute just retain
Object in physical memory.If value when this value starts compared to system adds more than 10%, then may have very serious
Memory overflow.
Write disk rate (Memory Pages per Second):It is measured as solving hard page mistake and from disk
Read page or page is write to the speed of disk.If this value is more than 1,000, then may occur internal memory due to excessive paging and let out
Leakage.
Quantity (Memory Page Read/sec) is read in paging:The Working Set of expression process are relative in physics
Deposit too big, be paged to disk.The quantity of read operation is shown in it, the page number not comprising each read operation.High numerical value
Show there is memory bottleneck.If relatively low page read operation and very high Physical Disk/%Disk Time and
Physical Disk/Avg.Disk Queue Length occur simultaneously, then should have disk bottleneck.If continued to increase
Queue length do not have along with reduction page read rate, then internal memory may not enough.
Preferably, in the above embodiments of the present application, the Data Collection of each part in step S11 creates computer system
Device, each data collector preserves a watch-list, and watch-list have recorded the performance counter associated with data collector
Counter identification in, the count value of the following any one or more performance counters of the data collector of memory collection:Patrol
Volume disk availability (LogicalDisk %Free Space), disk idleness (PhysicalDisk %Idle Time),
The average addressing speed of disk (PhysicalDisk Avg.Disk Sec/Read), the average writing speed of disk
(PhysicalDisk Avg.Disk Sec/Write), disk average queue length (PhysicalDisk Avg.Disk
Queue Length), caching quantity (Memory Cache Bytes), wherein, by the count value of any one performance counter
It is compared with corresponding monitoring threshold, the step of determining the performance number of memory in computer system includes:
Step S1731, logic magnetic disc availability is compared with the tenth monitoring threshold, obtains the tenth monitoring result;
Step S1732, disk idleness is compared with the 11st monitoring threshold, obtains the 11st monitoring result;
Step S1733, the disk addressing speed that is averaged is compared with the 12nd monitoring threshold, obtains the 12nd monitoring knot
Really;
Step S1734, the disk writing speed that is averaged is compared with the 13rd monitoring threshold, obtains the 13rd monitoring knot
Really;
Step S1735, disk average queue length is compared with the 14th monitoring threshold, obtains the 14th monitoring knot
Really;
Step S1736, caching quantity is compared with the 15th monitoring threshold, the 15th monitoring result is obtained;
Wherein, as the tenth monitoring result, the 11st monitoring result, the 12nd monitoring result, the 13rd monitoring result, the tenth
Any one or more monitoring results in four monitoring results, the 15th monitoring result are more than or equal to corresponding monitoring for count value
During threshold value, memory performance reduction is determined.
Specifically, by above-mentioned steps S1731 to step S1736, by the data collector of memory, respectively collection with
The numerical value of the performance counter of data collector association, by the comparison with monitoring threshold, determines the performance reduction of memory.
Wherein, because memory is used to store the program and data on processing server, therefore memory service condition and the speed of service
Bottleneck can greatly influence the overall performance of server.It can confirm to deposit using the combination of following performance counter in monitoring
The bottleneck of reservoir:
Logic magnetic disc availability (LogicalDisk %Free Space):It measures the available sky on selected logic magnetic disc
Between percentage.It note that if this value is less than 15%, then it represents that free space is not enough, and operating system can not store critical file.
One most direct solution is the more disk spaces of increase.
Disk idleness (PhysicalDisk %Idle Time):When it measures free time of the disk during the sampling interval
Between percentage.If this counter is less than 20%, then it represents that disk system is in full load condition.It is contemplated that by current disk
System conversion is the faster disk system of speed.
The average addressing speed of disk (PhysicalDisk Avg.Disk Sec/Read):It measures from disk and reads data
Average time (in seconds).If this numeral be more than 25 milliseconds (ms), mean from disk read data when disk
System is postponed.Most rational solution is that current disk system is replaced by into the faster disk system of speed.
The average writing speed of disk (PhysicalDisk Avg.Disk Sec/Write):It measures and writes data into magnetic
Average time (in seconds) needed for disk.If this numeral is more than 25ms, mean that disk system is sent out during write-in disk
Delay is given birth to.Possible solution it is that current disk system is replaced by the faster disk system of speed.
Disk average queue length (PhysicalDisk Avg.Disk Queue Length):It is represented currently etc.
Treat that hard disk is changed into available I/O operation number.If this value adds 2 more than the heart number of axle, then it represents that disk may be also bottleneck in itself.
Cache quantity (Memory Cache Bytes):It represents the amount of memory for file system cache.If this
Value is more than 200MB, then there may be disk bottleneck.
Preferably, in the above embodiments of the present application, as shown in Fig. 2 in step S17 by any one performance counter
Count value and corresponding monitoring threshold are compared, and after the performance number for determining each part in computer system, method is also
Including:
Step S19, determines that performance reduces type, and runnability according to the performance number of each part in computer system
Reduce the corresponding processing function module of type.
Specifically, by above-mentioned steps S19, the performance reduction type of all parts in computer system is carried out
Processing.Wherein, processing function module can include:Send prompting notification module, cleaning system background process, automatic diagnostic module
Deng.
By it is above-mentioned be directed to processor, internal memory, the combination of the counter of disk, collect performance counter data after, root
, being capable of quickly location processor, internal memory, the performance bottleneck of disk according to the comparison for collecting numerical value and threshold values.Meanwhile, according to bottleneck point
Propose solution.
Performance monitoring:It is a kind of systemic-function that Windows NT are provided.Windows NT are always integrated with all the time
Performance monitoring instrument, it provides the information about operating system current operating conditions.
Performance object (Performance Object):Monitored performance object, such as Processor, Process,
Memory, PhysicalDisk etc., equivalent to class (Class).
Performance counter (Performance Counter):The mode of performance objectivity energy information is described, equivalent to generic
Property.
Object instance (Object Instance):Same performance object may have multiple, and they are expressed as the object class
The different instances of type, equivalent to class example.
Embodiment 2
The embodiment of the present invention additionally provides a kind of device of monitoring system performance, as shown in figure 3, the device can include:
Creation module 21, setup module 23, read module 25, determining module 27.
Creation module 21, the data collector for creating each part in computer system, each data collector is protected
A watch-list is deposited, watch-list have recorded the counter identification of the performance counter associated with data collector.
Specifically, being each building block in computer system by creation module 21, a Data Collection is created
Device.The counter identification of performance counter corresponding with the building block in computer system is added in the data collector
Watch-list in.Data collector is obtained and the corresponding performance of watch-list Counter mark by calling watch-list
The count value of counter.Among practical application, the counting in watch-list can be changed according to business demand or actual conditions
Device identification entry.It is of course also possible to which by other means realize the function that watch-list is realized.
Setup module 23, for setting the monitoring threshold corresponding to each performance counter, constitutes monitoring threshold set.
Specifically, being the performance meter in list according to the watch-list in each data collector by setup module 23
Number device sets corresponding monitoring threshold, constitutes the set of monitoring threshold.Wherein, monitoring threshold is by for based on each performance
Number device judges the performance number of part corresponding with the performance counter.
Read module 25, the data collector for running any one or more parts, is read by data collector
The count value of the performance counter associated with data collector.
Specifically, by read module 25, being needed according to business or actual conditions, the data of one or more parts are run
Collector.Using data collector corresponding with all parts, the performance count associated with each data collector is obtained respectively
The count value of device.
Determining module 27, for the count value of any one performance counter and corresponding monitoring threshold to be compared,
Determine the performance number of each part in computer system.
Specifically, in determining module 27, the monitoring corresponding with each performance counter set by setup module 23
The count value for each performance counter that threshold value and step S15 are read, to the count value of any one performance counter and its
Corresponding monitoring threshold is compared, and determines the performance number of each part in computer system.
In above-mentioned creation module 21, setup module 23, read module 25, determining module 27, by performance counter
Select, combine, according to each part or the demand according to practical business of computer system, create data collector.
By data collector obtain performance counter count value, according to pre-set it is corresponding with each performance counter monitoring
Threshold value, fast and accurately judges the performance number of each part in computer system.
Among practical application, the present invention combines the contact between each module of Windows operating system bottom, is counted
The combination of device can be with Fast Discovery System bottleneck.By selecting, combinatorial enumeration device, and it is monitored and analyzed, it is quick fixed
Bit processor, internal memory, the performance bottleneck of disk.
Wherein, by running operation " performance monitor " in Windows management tools, after unlatching in " data collector "
Set up processor, internal memory, the data collector of disk.In processor, internal memory has been separately added into three data collectors of disk
Targetedly different counter.
Data after data are collected, are analyzed by operational data collection device, find out systematic function bottleneck point.
Preferably, in the above embodiments of the present application, in determining module 27, in addition to:First sub- monitoring modular 2711,
Two sub- monitoring modulars 2712, the 3rd sub- monitoring modular 2713, the 4th sub- monitoring modular 2714.
Wherein, the first sub- monitoring modular 2711, for processor average utilization to be compared with the first monitoring threshold,
Obtain the first monitoring result.
Second sub- monitoring modular 2712, for processor to be compared using the time with the second monitoring threshold, obtains
Two monitoring results.
3rd sub- monitoring modular 2713, for the processor break period to be compared with the 3rd monitoring threshold, obtains
Three monitoring results.
4th sub- monitoring modular 2714, for processor queue length to be compared with the 4th monitoring threshold, obtains
Four monitoring results.
Wherein, when any one in the first monitoring result, the second monitoring result, the 3rd monitoring result, the 4th monitoring result
When individual or multiple monitoring results are that count value is more than or equal to corresponding monitoring threshold, the performance reduction of processor is determined.
Specifically, in the above-mentioned first sub- monitoring modular 2711, the second sub- monitoring modular 2712, the 3rd sub- monitoring modular
2713rd, the 4th sub- monitoring modular 2714, by the data collector of processor, gathers the property associated with the data collector respectively
The numerical value of energy counter, by the comparison with monitoring threshold, determines the performance reduction of processor.Wherein, there is bottleneck in processor
It is probably because processor is in itself without enough abilities, or because application program inefficiency causes processor scarce capacity.
Processor bottleneck can be confirmed using the combination of following performance counter in monitoring:
Processor average utilization (Processor %Total Processor Time):The value measurement processor is used for
Perform non-NULL idle thread has been processor average utilization with percentage of time.If this value is more than 85%, show processing
Device performance is not enough, it may be necessary to be equipped with the faster processor of speed for server.
Processor utilizes time (Processor %User Time):The value measurement processor is spent in the user mode
Used percentage of time.If this value is higher, then it represents that server is busy in handling application program.Task manager can be opened
Check that application program takes cpu busy percentage.Close or optimization takes application program of the processor compared with large resource.
The processor break period (Processor %Interrupt Time):The value was measured in the specific phase in sampling interval
Between processor be used to receiving and handling time of hardware interrupts.If the value of this counter is more than 15%, then it represents that there may be
Hardware problem.
Processor queue length (System Processor Queue Length):It represents the line in processor queue
Number of passes.If this value is often than more than twice of CPU quantity, then it represents that the processor scarce capacity of server is, it is necessary to change many
Core processor.
Preferably, in the above embodiments of the present application, 27 in determining module in, in addition to:5th sub- monitoring modular 2721,
6th sub- monitoring modular 2722, the 7th sub- monitoring modular 2723, the 8th sub- monitoring modular 2724, the 9th sub- monitoring modular 2725.
Wherein, the 5th sub- monitoring modular 2721, for virtual memory utilization rate to be compared with the 5th monitoring threshold, is obtained
To the 5th monitoring result.
6th sub- monitoring modular 2722, for physical memory can be compared with ratio with the 6th monitoring threshold, is obtained
6th monitoring result.
7th sub- monitoring modular 2723, for nonpageable buffer pool size to be compared with the 7th monitoring threshold, is obtained
7th monitoring result.
8th sub- monitoring modular 2724, for write-in disk rate to be compared with the 8th monitoring threshold, obtains the 8th
Monitoring result.
9th sub- monitoring modular 2725, for paging reading quantity to be compared with the 9th monitoring threshold, obtains the 9th
Monitoring result.
Wherein, when the 5th monitoring result, the 6th monitoring result, the 7th monitoring result, the 8th monitoring result, the 9th monitoring knot
When any one or more monitoring results in fruit are more than or equal to corresponding monitoring threshold for count value, the performance drop of internal memory is determined
It is low.
Specifically, in the above-mentioned 5th sub- monitoring modular 2721, the 6th sub- monitoring modular 2722, the 7th sub- monitoring modular
2723rd, the 8th sub- monitoring modular 2724, the 9th sub- monitoring modular 2725, by the data collector of internal memory, collection is with being somebody's turn to do respectively
The numerical value of the performance counter of data collector association, by the comparison with monitoring threshold, determines the performance reduction of internal memory.Its
In, low memory is often as RAM deficiencies, memory overflow.Normal conditions internal memory is more, and magnetic disc i/o activity is fewer, therefore
Monitoring can use following counter to diagnose possible memory bottleneck.The combination of following performance counter can be used in monitoring
To confirm memory bottleneck:
Virtual memory utilization rate (Memory %Committed Bytes in Use):It measures Committed Bytes
With Commit Limit ratio, in other words, what it was measured is the virtual memory quantity in use.If this numerical value is more than
80%, then it represents that low memory.To this, most direct solution is the more internal memories of increase.
Physical memory can use ratio (Memory %Available Mbytes):Memory %Available Mbytes
It measures the physical memory amount (unit for Mbytes) available for operation process.If this value is less than the 5% of total physics RAM,
Mean low memory, paging activity may increase.It is the result that calculates recently and not average value to note this counter.
This problem is solved, more internal memories should be increased.
Nonpageable buffer pool size (Memory Pool Non-Paged Bytes):It measures the big of nonpageable buffer pool
Small (unit is byte).But as long as what is stored in this Installed System Memory region is that can not write disk and distribute just retain
Object in physical memory.If value when this value starts compared to system adds more than 10%, then may have very serious
Memory overflow.
Write disk rate (Memory Pages per Second):It is measured as solving hard page mistake and from disk
Read page or page is write to the speed of disk.If this value is more than 1,000, then may occur internal memory due to excessive paging and let out
Leakage.
Quantity (Memory Page Read/sec) is read in paging:The Working Set of expression process are relative in physics
Deposit too big, be paged to disk.The quantity of read operation is shown in it, the page number not comprising each read operation.High numerical value
Show there is memory bottleneck.If relatively low page read operation and very high Physical Disk/%Disk Time and
Physical Disk/Avg.Disk Queue Length occur simultaneously, then should have disk bottleneck.If continued to increase
Queue length do not have along with reduction page read rate, then internal memory may not enough.
Preferably, in the above embodiments of the present application, 27 in determining module in, in addition to:Tenth sub- monitoring modular 2731,
11st sub- monitoring modular 2732, the 12nd sub- monitoring modular 2733, the 13rd sub- of monitoring modular the 2734, the 14th monitoring mould
Block 2735, the 15th sub- monitoring modular 2736.
Wherein, the tenth sub- monitoring modular 2731, for logic magnetic disc availability to be compared with the tenth monitoring threshold, is obtained
To the tenth monitoring result.
11st sub- monitoring modular 2732, for disk idleness to be compared with the 11st monitoring threshold, obtains
11 monitoring results.
12nd sub- monitoring modular 2733, for the disk addressing speed that is averaged to be compared with the 12nd monitoring threshold,
Obtain the 12nd monitoring result.
13rd sub- monitoring modular 2734, for the disk writing speed that is averaged to be compared with the 13rd monitoring threshold,
Obtain the 13rd monitoring result.
14th sub- monitoring modular 2735, for disk average queue length to be compared with the 14th monitoring threshold,
Obtain the 14th monitoring result.
15th sub- monitoring modular 2736, for caching quantity to be compared with the 15th monitoring threshold, obtains the tenth
Five monitoring results.
Wherein, as the tenth monitoring result, the 11st monitoring result, the 12nd monitoring result, the 13rd monitoring result, the tenth
Any one or more monitoring results in four monitoring results, the 15th monitoring result are more than or equal to corresponding monitoring for count value
During threshold value, memory performance reduction is determined.
Specifically, in the above-mentioned tenth sub- monitoring modular 2731, the 11st sub- monitoring modular 2732, the 12nd sub- monitoring modular
2733rd, the 13rd sub- monitoring modular 2734, the 14th sub- monitoring modular 2735, the 15th sub- monitoring modular 2736, pass through storage
The data collector of device, the numerical value of performance counter that collection is associated with the data collector respectively, by with monitoring threshold
Compare, determine the performance reduction of memory.Wherein, because memory is used to store the program and data on processing server, because
The bottleneck of this memory service condition and the speed of service can greatly influence the overall performance of server.Under being used in monitoring
The combination of row performance counter confirms the bottleneck of memory:
Logic magnetic disc availability (LogicalDisk %Free Space):It measures the available sky on selected logic magnetic disc
Between percentage.It note that if this value is less than 15%, then it represents that free space is not enough, and operating system can not store critical file.
One most direct solution is the more disk spaces of increase.
Disk idleness (PhysicalDisk %Idle Time):When it measures free time of the disk during the sampling interval
Between percentage.If this counter is less than 20%, then it represents that disk system is in full load condition.It is contemplated that by current disk
System conversion is the faster disk system of speed.
The average addressing speed of disk (PhysicalDisk Avg.Disk Sec/Read):It measures from disk and reads data
Average time (in seconds).If this numeral be more than 25 milliseconds (ms), mean from disk read data when disk
System is postponed.Most rational solution is that current disk system is replaced by into the faster disk system of speed.
The average writing speed of disk (PhysicalDisk Avg.Disk Sec/Write):It measures and writes data into magnetic
Average time (in seconds) needed for disk.If this numeral is more than 25ms, mean that disk system is sent out during write-in disk
Delay is given birth to.Possible solution it is that current disk system is replaced by the faster disk system of speed.
Disk average queue length (PhysicalDisk Avg.Disk Queue Length):It is represented currently etc.
Treat that hard disk is changed into available I/O operation number.If this value adds 2 more than the heart number of axle, then it represents that disk may be also bottleneck in itself.
Cache quantity (Memory Cache Bytes):It represents the amount of memory for file system cache.If this
Value is more than 200MB, then there may be disk bottleneck.
Preferably, in the above embodiments of the present application, as shown in figure 4, also including in the device of above-mentioned monitoring system performance:
Processing module 29.
Wherein, processing module 29, determine that performance reduces class for the performance number according to each part in computer system
Type, and obtain the corresponding processing function module of performance reduction type.
Specifically, by above-mentioned processing module 29, the performance reduction type of all parts in computer system is entered
Row processing.Wherein, processing function module can include:Send prompting notification module, cleaning system background process, diagnose mould automatically
Block etc..
By it is above-mentioned be directed to processor, internal memory, the combination of the counter of disk, collect performance counter data after, root
, being capable of quickly location processor, internal memory, the performance bottleneck of disk according to the comparison for collecting numerical value and threshold values.Meanwhile, according to bottleneck point
Propose solution.
Performance monitoring:It is a kind of systemic-function that Windows NT are provided.Windows NT are always integrated with all the time
Performance monitoring instrument, it provides the information about operating system current operating conditions.
Performance object (Performance Object):Monitored performance object, such as Processor, Process,
Memory, PhysicalDisk etc., equivalent to class (Class).
Performance counter (Performance Counter):The mode of performance objectivity energy information is described, equivalent to generic
Property.
Object instance (Object Instance):Same performance object may have multiple, and they are expressed as the object class
The different instances of type, equivalent to class example.
It should be noted that for foregoing each method embodiment, in order to be briefly described, therefore it is all expressed as a series of
Combination of actions, but those skilled in the art should know, the present invention is not limited by described sequence of movement because
According to the present invention, some steps can be carried out sequentially or simultaneously using other.Secondly, those skilled in the art should also know
Know, embodiment described in this description belongs to preferred embodiment, involved action and module is not necessarily of the invention
It is necessary.
In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and does not have the portion being described in detail in some embodiment
Point, it may refer to the associated description of other embodiment.
, can be by another way in several embodiments provided herein, it should be understood that disclosed device
Realize.For example, device embodiment described above is only schematical, such as the division of described unit is only one kind
Division of logic function, can there is other dividing mode when actually realizing, such as multiple units or component can combine or can
To be integrated into another system, or some features can be ignored, or not perform.It is another, it is shown or discussed each other
Coupling direct-coupling or communication connection can be by some interfaces, the INDIRECT COUPLING or communication connection of device or unit,
Can be electrical or other forms.
The unit illustrated as separating component can be or may not be it is physically separate, it is aobvious as unit
The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can be selected to realize the mesh of this embodiment scheme according to the actual needs
's.
In addition, each functional unit in each embodiment of the invention can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units it is integrated in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.
If the integrated unit is realized using in the form of SFU software functional unit and as independent production marketing or used
When, it can be stored in a computer read/write memory medium.Understood based on such, technical scheme is substantially
The part contributed in other words to prior art or all or part of the technical scheme can be in the form of software products
Embody, the computer software product is stored in a storage medium, including some instructions are to cause a computer
Equipment (can be personal computer, mobile terminal, server or network equipment etc.) performs side described in each embodiment of the invention
The all or part of step of method.And foregoing storage medium includes:USB flash disk, read-only storage (ROM, Read-Only Memory),
Random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disc or CD etc. are various to be stored
The medium of program code.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, for the skill of this area
For art personnel, the present invention can have various modifications and variations.Within the spirit and principles of the invention, that is made any repaiies
Change, equivalent substitution, improvement etc., should be included in the scope of the protection.