CN105472009A - A method suitable for adaptive monitoring frequency of cloud platform resources - Google Patents

Abstract

The invention relates to the technical field of cloud platform monitoring, in particular to a method suitable for self-adaptive monitoring of cloud platform resources. Firstly, setting a monitoring item; the monitoring management terminal acquires monitoring item setting information; the monitoring agent reports data periodically according to the monitoring frequency; the monitoring management end analyzes data; if the threshold value is not reached, the monitoring management end calculates the next monitoring frequency of the monitoring item, and if the threshold value is reached, the next monitoring frequency of the monitoring item is the minimum monitoring frequency; returning the latest monitoring frequency to the monitoring agent; and finally, the monitoring agent sets the monitoring frequency to be the latest value. The method solves the problem that the current cloud platform cannot monitor the frequency in a self-adaptive manner when monitoring a plurality of virtual machines and computing nodes; the method can be applied to monitoring the cloud platform resources.

Description

A method suitable for adaptive monitoring frequency of cloud platform resources

technical field

The invention relates to the technical field of cloud platform monitoring, in particular to a method suitable for self-adaptive monitoring frequency of cloud platform resources.

Background technique

A general cloud computing platform has multiple computing nodes, and each computing node has multiple virtual machines, and some nodes with high performance even have dozens of virtual machines. The monitoring agent monitors these virtual machines and the computing nodes themselves need to consume some system resources. In most cases, computing nodes and virtual machines are in a normal state. When the customer wants to see, when the monitoring status of the virtual machine or physical machine is normal for a long time, there is no need to monitor it frequently. The following questions:

First, the more virtual machines there are, the more system resources will be occupied by the monitoring agent on the computing node, which will affect the normal operation of the virtual machines.

The second is that the monitoring frequency of conventional monitoring agents cannot be self-adaptive, and generally monitors for a long time according to a specific frequency given by the program.

The third is that it cannot meet the monitoring frequency requirements of customers.

Fourth, it is impossible to dynamically adjust the monitoring frequency according to the increase in the monitoring data of the monitoring item.

In order to solve the above problems, a method suitable for adaptive monitoring frequency of cloud platform resources is needed. The monitoring system itself adapts the monitoring frequency according to the monitoring state of the monitoring object.

Contents of the invention

The technical problem solved by the present invention is to provide a method suitable for self-adaptive monitoring frequency of cloud platform resources; to solve the problem that the monitoring agent monitoring frequency cannot be self-adaptive, the monitoring object is in a normal state, and the monitoring agent occupies more system resources, and cannot monitor according to the monitoring items. The monitoring data rises to dynamically adjust the monitoring frequency and other issues.

The technical scheme that the present invention solves the problems of the technologies described above is:

Described method comprises the steps:

Step 1: Set monitoring items;

Step 2: The monitoring management terminal obtains the monitoring item setting information;

Step 3: The monitoring agent regularly reports data according to the monitoring frequency;

Step 4: Monitor and manage the analysis data, if the threshold is not reached, then execute step 5; if the threshold is reached, execute step 6;

Step 5: The monitoring management terminal calculates the next monitoring frequency of the monitoring item; execute step 7;

Step 6: The next monitoring frequency of the monitoring item = the minimum monitoring frequency;

Step 7: Return the latest monitoring frequency to the monitoring agent;

Step 8: The monitoring agent sets the monitoring frequency to the latest value; go to step 2.

The setting of the monitoring item is to set the comparison mode, threshold, minimum monitoring frequency, maximum monitoring frequency and time interval of the monitoring item.

There are two commonly used comparison methods for the comparison methods of the monitoring items: greater than or equal to, less than or equal to;

The minimum monitoring frequency mentioned above refers to the lower limit of the adaptive monitoring frequency.

The maximum monitoring frequency mentioned above refers to the upper limit of the adaptive monitoring frequency.

The time interval mentioned refers to how long the continuous time does not reach the threshold before entering the maximum monitoring frequency. The user can set how many years, how many months, how many days, how many weeks or how many hours, and then the background of the program will automatically convert it into the smallest unit, usually The minimum monitoring frequency is 1 minute, then convert it into minutes, such as 1 week time = 7X24X60 minutes = 10080 minutes.

The analysis data of the monitoring management terminal is based on the latest monitoring item data reported, comparing whether the threshold is reached, comparing the latest monitoring value with the last monitoring value.

The monitoring management terminal calculates the monitoring frequency of the monitoring item next time, and the monitoring frequency of the next time=minimum monitoring frequency+((maximum monitoring frequency-minimum monitoring frequency)Xlog(normal state cumulative value-((increasing cumulative value/monitoring item Maximum value) X time interval))/log (time interval)), where X represents multiplication, and log represents a logarithmic function; if the calculation result of the next monitoring frequency is greater than the maximum monitoring frequency, then set the next monitoring frequency = maximum Monitoring frequency; if the calculation result of the next monitoring frequency is less than the minimum monitoring frequency, then set the next monitoring frequency = the minimum monitoring frequency; when the cumulative value of the normal state is larger, it will be closer to or reach the maximum monitoring frequency; when the cumulative value increases The bigger it is, the closer it is to or reaches the minimum monitoring frequency;

The accumulated value of the normal state=the accumulated value of the last normal state+1;

Described rising cumulative value, if the comparison mode is greater than or equal to, then the rising cumulative value=the last rising cumulative value+(the monitoring data of the latest monitoring item-the monitoring data of the old monitoring item); if the comparison mode is less than or equal, then the rising Accumulated value = last rising accumulative value + (monitoring data of the old monitoring item-monitoring data of the latest monitoring item);

The maximum value of the monitoring item is initially defaulted by the program to a maximum value, and when the monitoring data of the monitoring item is found to be greater than the maximum value, then set the maximum value=monitoring data of the monitoring item.

For the next "monitoring frequency=minimum monitoring frequency" of the monitoring item, set the next monitoring frequency as the minimum monitoring frequency; set the accumulated value of normal state to zero; set the accumulated value of rising to zero.

The present invention uses a method suitable for self-adaptive monitoring frequency of cloud platform resources, which solves the problem that monitoring agent monitoring frequency cannot be self-adaptive, the more monitoring objects are in a normal state, the more system resources occupied by monitoring agents, and the more system resources cannot be monitored according to monitoring items. The monitoring data rises to dynamically adjust the monitoring frequency and other issues. In the present invention, the monitoring agent actively reports the monitoring data, and dynamically adjusts the monitoring frequency according to the frequency returned by the monitoring management terminal, and the cloud platform monitoring agent in a general area. The present invention provides a calculation formula for adaptive frequency, and the cloud platform monitors and manages the general area.

Description of drawings

Below in conjunction with accompanying drawing, the present invention is further described:

Fig. 1 is a flowchart of the present invention;

Fig. 2 is a logical structure diagram of the present invention.

detailed description

There are multiple implementations of the present invention, and here a cloud platform is used as an example to illustrate one of the implementation methods, as shown in Figures 1 and 2, and the specific implementation process is as follows:

1. Set monitoring items;

/**

＊Set monitoring items

＊paramcompare comparison method

*paramthreshold threshold

*parammin minimum monitoring frequency

*parammax maximum monitoring frequency

*paramperiod time interval

*return boolean alarm evaluation and alarm task results

*/

public booleanvalationAndAction(Stringcompare, Longthreshold, intmin, intmax, intperiod){

return Rule.setThreshold (compare, threshold, min, max, period);

}

2. The monitoring management terminal obtains the monitoring item setting information;

/**

＊The monitoring management terminal obtains the monitoring item setting information

*paramobjectId monitoring item Id

＊returnRule fetches monitoring item setting information

*/

public Rule getRule(stringobjectId){

return Rules.getRule (objectId);

}

3. The monitoring agent regularly reports data according to the monitoring frequency;

/**

＊The monitoring agent regularly reports data according to the monitoring frequency

＊returnvoid no return result

*/

public void requestMonitorData(){

return Service.requestMonitorData ();

}

4. The monitoring management terminal analyzes the data, and if the threshold is not reached, the process of calculating the monitoring frequency is executed; if the threshold is reached, the process of setting the minimum monitoring frequency is executed

/**

＊Monitoring management terminal analysis data

*paramdata monitoring data object

* return boolean analysis result

*/

public boolean analysis (DataModel data ) {

return ThresholdRule . analysis(data);

}

5. The monitoring management terminal calculates the next monitoring frequency of the monitoring item; executes the process of returning to the monitoring frequency;

/**

＊The monitoring management terminal calculates the next monitoring frequency of the monitoring items

*paramdata monitoring data object

＊returnLong new monitoring frequency

*/

publicLongcaculate (DataModel data ) {

return ThresholdRule.caculate (data);

}

6. Set the next monitoring frequency of the monitoring item to the minimum monitoring frequency;

/**

＊Set the next monitoring frequency of the monitoring item to the minimum monitoring frequency;

＊returnvoid no return result

*/

public void setFrequency( ) {

this.frequency = this.min ;

}

7. Return the latest monitoring frequency to the monitoring agent;

/**

＊Return the latest monitoring frequency to the monitoring agent

＊paramfrequency latest monitoring frequency

＊returnvoid no return result

*/

public void sendResponse ( Longfrequency ) {

Service . sendResponse(frequency);

}

8. The monitoring agent sets the monitoring frequency to the latest value; re-executes the reporting of the monitoring data of the monitoring items.

/**

＊The monitoring agent sets the monitoring frequency to the latest value

＊paramfrequency latest monitoring frequency

＊returnvoid no return result

*/

public void setFrequency ( Longfrequency ) {

Service.setFrequency (frequency);

}

The whole process is over.

Claims (7)

1. be applicable to a method for cloud platform resource adaptive process monitoring frequency, it is characterized in that: described method comprises the steps:

Step 1: monitored item is set;

Step 2: monitoring management end obtains monitored item configuration information;

Step 3: monitoring agent is according to the regular reported data of monitoring frequency;

Step 4: monitoring management end analyzes data, if do not reach threshold value, performs step 5; If reach threshold value, perform step 6;

Step 5: monitoring management end calculates the monitoring frequency of monitored item next time; Perform step 7;

Step 6: the monitoring frequency=minimum monitoring frequency of monitored item next time;

Step 7: return up-to-date monitoring frequency to monitoring agent;

Step 8: it is last look that monitoring agent arranges monitoring frequency; Perform step 2.

2. method according to claim 1, is characterized in that: the described monitored item that arranges arranges the manner of comparison of monitored item, threshold value, minimum monitoring frequency, maximum monitoring frequency and the time interval.

The manner of comparison of described monitored item, conventional manner of comparison has two kinds: be more than or equal to, be less than or equal to;

Described minimum monitoring frequency, minimum monitoring frequency refers to the lower limit of adaptive process monitoring frequency.

Described maximum monitoring frequency, maximum monitoring frequency refers to the upper limit of adaptive process monitoring frequency.

The described time interval, refer to that how long not reaching threshold value just enters maximum monitoring frequency continuously, user can arrange how many years, Duo Shaoyue, Duo Shao, how many week or how many hours, then program backstage changes into least unit automatically, usual monitoring frequency is minimum is 1 minute, so just change into a minute unit, as 1 time-of-week=7X24X60 minute=10080 minutes.

3. method according to claim 1, is characterized in that: it is up-to-date monitored item data that basis reports that described monitoring management end analyzes data, more whether reaches threshold value, more up-to-date monitoring value and last monitoring value.

4. method according to claim 2, is characterized in that: it is up-to-date monitored item data that basis reports that described monitoring management end analyzes data, more whether reaches threshold value, more up-to-date monitoring value and last monitoring value.

5. the method according to any one of Claims 1-4, it is characterized in that: described monitoring management end calculates the monitoring frequency of monitored item next time, monitoring frequency=minimum monitoring frequency+((maximum monitoring frequency-minimum monitoring frequency) Xlog (normal condition accumulated value-((rising accumulated value/monitored item maximum) the X time interval))/log (time interval)) of next time, wherein X represents and is multiplied by, and log represents logarithmic function; If the monitoring frequency computation part result of next time is greater than maximum monitoring frequency, then the monitoring frequency=maximum monitoring frequency of next time is set; If the monitoring frequency computation part result of next time is less than minimum monitoring frequency, then the monitoring frequency=minimum monitoring frequency of next time is set; When normal condition accumulated value is larger, so more close or reach maximum monitoring frequency; When rising accumulated value is larger, so more close or reach minimum monitoring frequency;

Described normal condition accumulated value=last time normal condition accumulated value+1;

Described rising accumulated value, if manner of comparison is more than or equal to, the accumulated value that rose accumulated value=last time of so rising+(monitor data of the monitor data-old monitored item of up-to-date monitored item); If manner of comparison is less than or equal to, the accumulated value that rose accumulated value=last time of so rising+(monitor data of the monitor data-up-to-date monitored item of old monitored item);

Described monitored item maximum, gives tacit consent to a maximum by program at the beginning, when finding that the monitor data of monitored item is greater than maximum, then arranges the monitor data of maximum=monitored item.

6. the method according to any one of Claims 1-4, is characterized in that: " the monitoring frequency=minimum monitoring frequency " of described monitored item next time, and the monitoring frequency arranging next time is minimum monitoring frequency; Arranging normal condition accumulated value is zero; Arranging rising accumulated value is zero.

7. method according to claim 5, is characterized in that: " the monitoring frequency=minimum monitoring frequency " of described monitored item next time, and the monitoring frequency arranging next time is minimum monitoring frequency; Arranging normal condition accumulated value is zero; Arranging rising accumulated value is zero.