WO2014063557A1 - 网元负载不均检测处理方法、装置及其系统 - Google Patents

网元负载不均检测处理方法、装置及其系统 Download PDF

Info

Publication number
WO2014063557A1
WO2014063557A1 PCT/CN2013/084512 CN2013084512W WO2014063557A1 WO 2014063557 A1 WO2014063557 A1 WO 2014063557A1 CN 2013084512 W CN2013084512 W CN 2013084512W WO 2014063557 A1 WO2014063557 A1 WO 2014063557A1
Authority
WO
WIPO (PCT)
Prior art keywords
network element
load
threshold
detection
network
Prior art date
Application number
PCT/CN2013/084512
Other languages
English (en)
French (fr)
Inventor
姜春鹤
刘峰
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Priority to EP13849007.3A priority Critical patent/EP2899918A4/en
Priority to US14/437,337 priority patent/US9674065B2/en
Publication of WO2014063557A1 publication Critical patent/WO2014063557A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0876Network utilisation, e.g. volume of load or congestion level
    • H04L43/0888Throughput
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0805Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
    • H04L43/0817Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06Management of faults, events, alarms or notifications
    • H04L41/0681Configuration of triggering conditions
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06Management of faults, events, alarms or notifications
    • H04L41/069Management of faults, events, alarms or notifications using logs of notifications; Post-processing of notifications
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/16Threshold monitoring
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/11Identifying congestion
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08Configuration management of networks or network elements
    • H04L41/0803Configuration setting
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network
    • H04L67/1001Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1034Reaction to server failures by a load balancer

Definitions

  • Network element load unevenness detection processing method device and system thereof
  • the present invention relates to the field of network management technologies, and in particular, to a network element load unevenness detection processing method, apparatus, and system thereof. Background technique
  • the network element management system (centralized network management system) can manage multiple network element devices at the same time. This function has been increasingly favored and valued by operators. Users can see the operation of each network element device in real time through the network element management system. Status and health indicators, so network operation and maintenance personnel can perform timely maintenance on the NE devices that have problems, which can improve the reliability of network devices.
  • the existing network element management system architecture deploys various processing units and network element agent units in the network element device, and deploys the network element manager unit and the alert unit in the network element management system.
  • the processing unit sends the corresponding data information to the network element agent unit after the network service occurs, and the network element agent unit generates an important network performance indicator according to the data information reported by each processing unit and saves the information to the network element device buffer pool.
  • the network element agent unit reports the indicator in the network element indicator buffer pool to the network element manager unit, and the network element manager unit performs the received network element device indicator. After being properly processed, it is saved in the network element M indicator buffer pool or the network element N indicator buffer pool.
  • the network element manager unit periodically detects the network element device in the network element M indicator buffer pool and the network element N indicator buffer pool. Whether there is an abnormality in the indicator (for example, the load is too heavy). If there is an abnormality, the abnormality information is reported to the warning unit in time to alert the user. For example, the warning device such as the sound and light alarm prompts the user to deal with the abnormal problem as soon as possible.
  • the warning device such as the sound and light alarm prompts the user to deal with the abnormal problem as soon as possible.
  • this type of processing can meet the operation and maintenance requirements of the most basic network element equipment of the operation and maintenance personnel.
  • this method is only for the detection of the separate network element device, and the warning occurs only after the network element device actually has a problem. In the face of increasing user attention to equipment automation, multi-network element association analysis and fault prevention requirements, it is obviously not enough.
  • the purpose of the embodiments of the present invention is to provide a network element load unevenness detection, in order to solve the problem that the network element management system cannot perform the automatic difference analysis of the network element device indicators to find out whether the network element devices have security risks. Processing method, device and system thereof.
  • a method for detecting a network element load unevenness including:
  • the NE performance indicators reported by the NE devices are differentiated according to the preset detection policy, and the analysis results are saved in the network indicator buffer pool.
  • the differential analysis result is obtained from the network indicator buffer pool periodically, and it is judged whether there is an abnormal load imbalance between the network element devices. When it is judged that an abnormality occurs, a warning message is issued.
  • the detection strategy includes a load percentage threshold policy, a constraint absolute value threshold policy, and a load ratio policy.
  • the alert message includes current indicator information, threshold information, and load ratio information.
  • the network element performance indicator reported by each network element device is differentiated and analyzed according to a preset detection policy, including:
  • the packet determines whether the network element performance index of each network element device managed by the network element management system is complete;
  • the load mean deviation is calculated according to the network element performance index and the load ratio of each network element device, and the load mean deviation is used as a detection reference for whether the network element network element performance indicator is overloaded, where the load mean deviation P Calculated as follows:
  • Ci the current value of each network element performance indicator
  • Di the load ratio of each network element device
  • Vx (Cx/Dx-P)*100/P, where Vx is the percentage of load deviation of each network element device, Cx is the current value of each network element performance indicator, and Dx is the load ratio of each network element device, P The load mean deviation of each network element device;
  • the load deviation percentage of each network element device is compared with the preset load percentage threshold. If the threshold is exceeded, the network element performance indicator of the corresponding network element device is abnormal. Otherwise, the corresponding network element device is identified. The NE performance indicator is normal.
  • the load percentage threshold includes a positive threshold and a negative threshold. If Vx is a positive value, it is compared with a positive threshold of a load percentage threshold. If Vx is a negative value, it is compared with a load percentage. The negative threshold of the threshold is compared.
  • an alert message is sent, including:
  • a network element load unevenness detecting and processing device includes:
  • the load-unbalanced policy configuration unit is configured to parameterize the detection policy for performing differential analysis on the network element performance indicators of each network element device;
  • the network element performance indicator buffer pool is configured to store the network element performance indicators reported by the network element devices.
  • the load unevenness detection unit is configured to differentiate the network element performance indicators reported by the network element devices according to a preset detection policy.
  • the analysis results are saved in the network indicator buffer pool.
  • the network indicator buffer pool is configured to store the analysis results of the network element performance indicators of each network element device after differential analysis.
  • the detection strategy includes a load percentage threshold policy, a constraint absolute value threshold policy, and a load ratio policy.
  • a network element load unevenness detection processing system includes:
  • the NE device is configured to report the performance indicators of the NEs.
  • the network element load unevenness detection processing device is configured to perform differential analysis on the network element performance indicators reported by each network element device according to a preset detection policy, and save the analysis result in the network indicator buffer pool;
  • the network element management system is configured to periodically obtain the differential analysis result from the network indicator buffer pool of the network element load unevenness detection processing device, and determine whether the load unevenness abnormality occurs between the network element devices according to the determination, when determining When an exception occurs, an alert message is issued.
  • the detection strategy includes a load percentage threshold policy, a constraint absolute value threshold policy, and a load ratio policy.
  • the alert message includes current indicator information, threshold information, and load ratio information.
  • the network element load unevenness detecting processing device is configured to determine, according to the indicator information in the network element performance indicator buffer pool, that the network element performance indicator of each network element device managed by the network element management system is Complete;
  • the load mean deviation is calculated according to the network element performance index and the load ratio of each network element device, and the load mean deviation is used as a detection reference for whether the network element network element performance indicator is overloaded, where the load mean deviation P Calculated as follows:
  • Ci the current value of each network element performance indicator
  • Di the load ratio of each network element device
  • Vx (Cx/Dx-P)*100/P, where Vx is the percentage of load deviation of each network element device, Cx is the current value of each network element performance indicator, and Dx is the load ratio of each network element device, P The load mean deviation of each network element device;
  • the load deviation percentage of each network element device is compared with the preset load percentage threshold. If the threshold is exceeded, the network element performance indicator of the corresponding network element device is abnormal. Otherwise, the corresponding network element device is identified. The NE performance indicator is normal.
  • the load percentage threshold includes a positive threshold and a negative threshold. If Vx is a positive value, it is compared with a positive threshold of a load percentage threshold. If Vx is a negative value, it is compared with a load percentage. The negative threshold of the threshold is compared.
  • the network element management system is configured to detect whether a warning has been issued and the warning has not been eliminated, if yes, no redundant network element device overload warning is issued; otherwise, a new network element device overload warning is issued, which carries The current indicator information, threshold information, and load ratio information.
  • the embodiment of the present invention adopts load mean deviation calculation, load deviation percentage calculation, and threshold comparison technology, and can be used in various network element devices.
  • load mean deviation calculation load deviation percentage calculation
  • threshold comparison technology threshold comparison technology
  • FIG. 1 is a schematic diagram of an existing network element overload processing framework
  • FIG. 2 is a schematic flow chart of a method for processing a network element load unevenness detection according to an embodiment of the present invention
  • FIG. 3 is a schematic diagram of networking for detecting network load unevenness according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of a network element load unevenness detecting and processing system according to an embodiment of the present invention.
  • FIG. 5 is a flowchart of a method for detecting a network element load unevenness detection according to an embodiment of the present invention. detailed description
  • a method for detecting a network element load unevenness detection includes the following steps:
  • S100 Perform differential analysis on the network element performance indicators reported by each network element device according to a preset detection policy, and save the analysis result to the network indicator buffer pool;
  • S200 Obtain the differential analysis result from the network indicator buffer pool periodically, and judge whether there is an abnormal load imbalance between the network element devices according to the determination, and when it is determined that an abnormality occurs, a warning message is sent.
  • the detection policy includes a load percentage threshold policy, a constraint absolute value threshold policy, and a load ratio policy.
  • the alert message includes current indicator information, threshold information, and load ratio information.
  • the network element performance indicator includes two application scenarios:
  • One is to detect load balancing of similar network elements, for example, detecting load balancing of each MSC network element in a MSC (Mobile Switching Center) pool (POOL), and detecting SGSN (Serving GPRS support Node, serving GPRS support node)
  • the load balancing of each SGSN network element in the POOL may be a service indicator.
  • the metric detected by the SGSN POOL may be the number of attached users, the average number of sessions in the active state, and the Gn/Gp port traffic.
  • the other is to detect load balancing of different network elements, such as detecting SGSN and GGSN (Gateway
  • the detected indicators are generally hardware-related common indicators, such as CPU usage, memory usage, network traffic, etc., and may also be logically associated service indicators, such as The association between the number of SGSN-attached users and the number of GGSN bearers is shown in Figure 3.
  • the network element provides the corresponding index value.
  • the network element load-uneven detection processing device functions as a separate device or as a network element. A separate device deployment in the management system that performs load balancing detection on a portion of the network elements that the user cares about.
  • the step of performing differentiated analysis on the network element performance indicator reported by each network element device according to the preset detection policy includes:
  • step A Determine whether the network element performance indicators of the network element devices managed by the network element management system are complete according to the indicator information in the network element performance indicator buffer pool. If complete, perform step B, otherwise perform step C;
  • step B Determine whether the network element performance indicators of the network element devices have exceeded the preset constraint absolute value threshold, and if yes, perform step D, otherwise perform step C;
  • the load mean deviation is used as a detection criterion for whether the network element performance indicator of the network element is overloaded.
  • the calculation formula of the load mean deviation P is as follows:
  • Ci the current value of each network element performance indicator
  • Di the load ratio of each network element device
  • Vx (Cx/Dx-P)*100/P, where Vx is the percentage of load deviation of each network element device, Cx is the current value of each network element performance indicator, and Dx is the load ratio of each network element device, P The load mean deviation of each network element device;
  • step F Compare the load deviation percentage of each network element device with a preset load percentage threshold. If the threshold is exceeded, perform step G, otherwise perform step H;
  • the NE performance indicator of the corresponding NE device is abnormal.
  • the NE performance indicator of the corresponding NE device is normal.
  • the load percentage threshold includes a positive threshold and a negative threshold. If Vx is a positive value, it is compared with a positive threshold of a load percentage threshold. If Vx is a negative value, Compare this to the negative threshold of the load percentage threshold.
  • the step of issuing the warning message includes:
  • step a detecting whether a warning has been issued before and the warning has not been eliminated, if yes, proceed to step b, otherwise perform step c;
  • the method further includes: when the judgment is normal, detecting whether a warning has been issued before and the warning has not been eliminated, and if yes, canceling the overload warning of the network element device, otherwise not doing Any processing.
  • the network element load unevenness detecting processing method provided by the embodiment of the present invention, the method implemented by the network element load unevenness detecting processing device mainly includes the following steps:
  • Step 1 Perform differential analysis on the network element performance indicators reported by each network element device according to the set detection policy, and save the analysis result after the differential analysis to the network indicator buffer pool, where the detection strategy includes a load percentage threshold. , Constrained absolute value threshold, load ratio.
  • the first part is the load percentage threshold, including the positive threshold and the negative threshold, and both the positive threshold and the negative threshold can be configured with multiple thresholds, which are configured to indicate what level of warning is required for the load unevenness indicator exceeding what threshold;
  • the second part is the constraint absolute value threshold, which is configured to indicate how many indicators above the base need to be detected for load unevenness, because it is acceptable to have a large difference in load of each network element when the amount of data to be processed is small;
  • the third part is the load ratio, which is configured to indicate how the processing power of each network element is. Because each network element has different processing capabilities due to its different software/hardware configurations, it is not the same ratio of 1:1.
  • Step 2 Send the differential analysis result of each network element device in the network indicator buffer pool to the network element manager unit of the network element management system, and the network element manager unit finds that the result is abnormal, and sends a warning message to the warning unit, and the warning information is sent. It includes current indicator information, threshold information, and load ratio.
  • the first part is the current indicator information, so that the operation and maintenance personnel can know the current load status of the current network element device in real time;
  • the second part is the current threshold information, so that the operation and maintenance personnel can understand the threshold configuration information when the warning message is sent. Because the threshold configuration information may be modified, the related information needs to be matched.
  • the third part is the load ratio information, so that the operation and maintenance personnel can understand the load ratio information when the warning message is sent, because the load ratio information can also be adjusted.
  • the embodiment of the present invention further provides a network element load unevenness detecting and processing device.
  • the method includes:
  • the load unevenness policy configuration unit 32 is configured to perform parameter configuration of the detection strategy for performing differential analysis on the network element performance indicators of each network element device;
  • the network element performance indicator buffer pool is configured to store the network element performance indicators reported by the network element devices.
  • the load unevenness detecting unit 31 is configured to perform the difference according to the preset detection policy for the network element performance indicators reported by the network element devices.
  • the analysis is performed, and the analysis result is saved in the network indicator buffer pool.
  • the network indicator buffer pool 33 is configured to store the analysis result of the network element performance index of each network element device after the differential analysis.
  • the network element performance indicator buffer pool includes a first network element M indicator buffer pool 34 and a first network element N indicator buffer pool 35, where:
  • the first network element M indicator buffer pool 34 is configured to store performance indicators of the network element M, and is used as a material for differential analysis of performance indicators of each network element device;
  • the first network element N indicator buffer pool 35 is configured to store the performance index of the network element N, and is used as a material for differential analysis of each network element device performance indicator;
  • the detection strategy includes a load percentage threshold policy, a constrained absolute value threshold policy, and a load proportional policy.
  • the load unevenness detecting unit 31, the load unevenness policy configuring unit 32, the network indicator buffer pool 33, the first network element M indicator buffer pool 34, and the first network element N indicator buffer pool 35 It can be implemented by a Central Processing Unit (CPU), a Digital Signal Processor (DSP), or a Field-Programmable Gate Array (FPGA).
  • CPU Central Processing Unit
  • DSP Digital Signal Processor
  • FPGA Field-Programmable Gate Array
  • an embodiment of the present invention provides a network element load unevenness detection processing system.
  • a network element load unevenness detection processing system include:
  • the NE device 10 is configured to report the performance indicators of the NEs.
  • the network element load unevenness detecting and processing device 30 is configured to perform differential analysis on the network element performance indicators reported by the network element devices according to the preset detection policy, and save the analysis result in the network indicator buffer pool;
  • the network element management system 20 is configured to periodically obtain the difference analysis result from the network indicator buffer pool of the network element load unevenness detecting processing device 30, and determine whether the load unevenness abnormality occurs between the network element devices according to the determination. When it is judged that an abnormality has occurred, a warning message is issued.
  • the detection strategy includes a load percentage threshold policy, a constraint absolute value threshold policy, and a load ratio policy.
  • the alert message includes current indicator information, threshold information, and load ratio information.
  • the network element load unevenness detecting processing device 30 is configured to determine, according to the indicator information in the network element performance indicator buffer pool, whether the network element performance indicators of the network element devices managed by the network element management system are complete;
  • the load mean deviation is calculated according to the network element performance index and the load ratio of each network element device, and the load mean deviation is used as a detection criterion for whether the network element network element performance indicator is overloaded, wherein the load mean deviation P is calculated.
  • the formula is as follows:
  • Ci the current value of each network element performance indicator
  • Di the load ratio of each network element device
  • Vx (Cx/Dx-P)*100/P, where Vx is the percentage of load deviation of each network element device, Cx is the current value of each network element performance indicator, and Dx is the load ratio of each network element device, P For each network element The load mean deviation of the equipment;
  • the ratio of the load deviation of each network element device is compared with the preset load percentage threshold. If the threshold is exceeded, go to step G. Otherwise, go to step H.
  • the NE performance indicator of the corresponding NE device is abnormal.
  • the NE performance indicator of the corresponding NE device is normal.
  • the load percentage threshold includes a positive threshold and a negative threshold. If Vx is a positive value, it is compared with a positive threshold of a load percentage threshold. If Vx is a negative value, it is negative with a load percentage threshold. Thresholds are compared.
  • the network element management system 20 is configured to detect whether a warning has been issued previously and the warning has not been eliminated, if the redundant network element device overload warning is no longer issued; otherwise, a new network element device overload warning is issued, where the carrying Current indicator information, threshold information, and load ratio information.
  • the network element load unevenness detecting processing device 30 for implementing the network element load balancing detection processing method shown in FIG. 2, as shown in FIG. 4, includes: interconnected load unevenness The detecting unit 31 and the load unevenness policy configuring unit 32, the interconnected load unevenness detecting unit 31 and the network indicator buffer pool 33, the interconnected network indicator buffer pool 33, and the first network element M indicator buffer pool 34 and The first network element N indicator buffer pool 35.
  • the load unevenness detecting unit 31 is configured to perform differential load balancing analysis on the network element performance index of the network element device reported by the network element device according to the configured detection policy, and save the differentiated analysis result to the network indicator buffer pool 33. in.
  • the load unevenness policy configuration unit 32 is configured to configure and store the detection policy parameters of the load unevenness.
  • the network indicator buffer pool 33 is configured to save the load balancing detection result of each network element device subjected to the differential analysis, and send the differential analysis result to the network element management system 20.
  • the first network element M indicator buffer pool 34 is configured to save the load balancing detection result of the M network element device that is subjected to differential analysis.
  • the first network element N indicator buffer pool 35 is configured to save the load balancing detection result of the N network element device subjected to the differential analysis.
  • the network element device 10 connected to the load unevenness detecting unit 31 includes: a processing unit 11, a network element, a proxy unit 12, and a network element device index buffer pool 13.
  • the processing unit 11 is configured to detect an operating state of the network element device, and report corresponding data information to the network element agent unit 12 when a service occurs on the network.
  • the network element agent 12 is connected to the processing unit 11 and the load unevenness detecting unit 31, and configured to receive the data information reported by the processing unit 11 and extract the network element performance indicator into the network element device buffer pool of the network element device. 13, and when appropriate (timed), the network element performance index of the network element device is reported to the load unevenness detecting unit 31 for performance index differentiation analysis, and the original network element performance network element performance index is also reported to
  • the network element management system 20 is configured to analyze network element device indicators of individual network elements.
  • the network element device buffer pool 13 is connected to the network element proxy unit 12 and configured to cache performance indicators of the network element device.
  • the network element management system 20 connected to the network indicator buffer pool 33 includes: a network element manager unit 21, a second network element M indicator buffer pool 22, a second network element N indicator buffer pool 23, and a warning unit 24.
  • the network element manager unit 21 is connected to the network element agent 12 and the network indicator buffer pool 33, and configured to receive the network element performance indicator of the network element device of the separate network element sent by the network element agent 12, and cached to The second network element M indicator buffer pool and the second network element N indicator buffer pool are analyzed and processed; and further configured to obtain differentiated load balancing of each network element device from the network indicator buffer pool 33 The detection result is analyzed and processed, and if an abnormality is found, an alert is issued to the warning unit 24.
  • the second network element M indicator buffer pool 22 is connected to the network element manager unit 21 and configured to cache the network element device indicators of the network element M.
  • the second network element N indicator buffer pool 23 is connected to the network element manager unit 21 and configured as a network.
  • the NE device indicator of the meta N is cached.
  • the alert unit 24 is connected to the network element manager unit 21 and configured to display and alert the alert message sent by the network element manager unit 21 to the alert message.
  • the network element manager unit 21, the second network element M indicator buffer pool 22, the second network element N indicator buffer pool 23, and the alert unit 24 may be implemented by a CPU, a DSP, or an FPGA.
  • FIG. 5 is a flowchart of a method for detecting a network element load unevenness detecting process according to an embodiment of the present invention. The method mainly includes the following steps:
  • step 401 the network element proxy unit 12 reports the network element device performance indicator of the network element to the load unevenness detecting unit 31, and the indicator must include a valid performance indicator value.
  • Step 402 The load unevenness detecting unit 31 detects whether the reported network element device indicator is complete according to the policy, that is, whether the network element monitored by the policy has reported the performance indicator. If yes, step 403 is performed, and if no, the process ends.
  • Step 403 The load unevenness detecting unit 31 detects whether the reported network element device metrics exceed the constraint absolute value threshold required by the policy according to the policy. If yes, step 404 is performed, and if no, the process ends.
  • Step 405 The load unevenness detecting unit 31 checks whether the network element device indicators have been detected. If yes, the process ends. If no, step 406 is performed.
  • Step 406 the load unevenness detecting unit 31 takes out an undetected network element device indicator.
  • Step 408 The load unevenness detecting unit 31 detects whether the percentage of the load deviation of the network element exceeds the load percentage threshold of the policy configuration. If yes, step 409 is performed, and if no, step 411 is performed.
  • Step 409 the load unevenness detecting unit 31 identifies the network element device indicator of the network element as an abnormality and puts it into the network indicator buffer pool 33, and repeats step 405.
  • Step 410 The load unevenness detecting unit 31 performs aging processing on the network element device index in the network indicator buffer pool 33.
  • Step 411 the load unevenness detecting unit 31 identifies the network element device indicator of the network element as normal and puts it into the network indicator buffer pool 33, and repeats step 405.
  • the process includes:
  • step 412 the network element manager unit 21 extracts the differential load balancing detection result of the network element device from the network indicator buffer pool 33 and performs analysis processing. If an abnormality is found, the process proceeds to step 413; if it is normal, the process proceeds to step 414. ;
  • Step 413 The network element manager unit 21 extracts the network element device indicator from the network indicator buffer pool 33, detects whether a corresponding alert has been sent and the alert has not been eliminated, and if yes, the process ends, and if not, executes Step 415.
  • Step 414 the network element manager unit 21 extracts the network element device indicator from the network indicator buffer pool 33, detects whether a corresponding alert has been sent and the alert has not been eliminated, and if yes, step 416 is performed, and if not, then The process ends.
  • step 415 the network element manager unit 21 sends a new network element device load unevenness warning message to the alert unit 24, and the alert message carries current indicator information, threshold information, and load ratio information.
  • step 416 the network element manager unit 21 sends an alert removal message to the alert unit 24 to eliminate the network device load imbalance warning.
  • a similar indicator in the element performs load unevenness detection to automatically discover whether there is a problem of uneven load of the logical unit in the network element.
  • the present invention has other various embodiments, for example, the load unevenness detecting unit is not directly connected to the network element manager unit through the network element agent unit, and the like.
  • the load unevenness detecting unit is not directly connected to the network element manager unit through the network element agent unit, and the like.
  • various changes and modifications can be made by those skilled in the art without departing from the spirit and scope of the invention, and the corresponding changes and modifications should fall within the scope of the present invention. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Debugging And Monitoring (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Computer And Data Communications (AREA)

Abstract

本发明公开了一种网元负载不均检测处理方法、装置及系统,其中,所述方法包括:对各网元设备上报的网元性能指标根据预设的检测策略进行差异化分析,并将分析结果保存到网络指标缓冲池中;定期从网络指标缓冲池中获取差异化分析结果,并据此判断各网元设备之间是否出现负载不均异常,当判断得出出现异常时,则发出警示消息。本发明采用负载均值偏差计算、负载偏差百分比计算以及与门限比对技术,能够在多种网元设备共管的场景下及时发现某些网元设备负载不均的问题,从而通过发出警示消息第一时间通知用户及时消除网元设备可能存在的安全隐患,以提升网络通讯设备的可靠性、可维护性和稳定性。

Description

网元负载不均检测处理方法、 装置及其系统 技术领域
本发明涉及网络管理技术领域, 具体而言, 尤其涉及一种网元负载不 均检测处理方法、 装置及其系统。 背景技术
在网络通信技术领域, 随着信息通信量的与日倶增, 用户对高稳定性 和高性能的通信产品的需求越来越大, 与此同时, 其对通信设备的可靠性 也就有了更高的要求。
目前, 网元管理系统(集中网管) 能够同时管理多个网元设备, 这种 功能已经越来越受到运营商的青睐和重视, 用户通过网元管理系统可以实 时看到各个网元设备的运行状态和健康指标, 这样网络操作维护人员可以 对出现问题的网元设备及时进行维护, 从而可以提高网络设备的可靠性。
如图 1 所示, 现有的网元管理系统架构是在网元设备中部署各种处理 单元和网元代理者单元, 在网元管理系统中部署网元管理者单元和警示单 元。 其中, 处理单元发生网络业务后会向网元代理者单元上报对应的数据 信息, 网元代理者单元根据各处理单元上报的数据信息生成重要的网络性 能指标并保存到网元设备指标緩冲池中, 在适当的时候(定时粒度), 网元 代理者单元会把网元设备指标緩冲池中的指标上报给网元管理者单元, 网 元管理者单元把接收到的网元设备指标进行适当的处理后保存到网元 M指 标緩冲池或者网元 N指标緩冲池中, 网元管理者单元定期检测网元 M指标 緩冲池和网元 N指标緩冲池中的网元设备指标是否存在异常(例如负载过 重),如果存在异常,则及时把异常信息上报给警示单元以对用户进行警示, 例如通过声光告警等警示设备提示用户尽快处理异常问题。 通常情况下, 这种处理已经可以满足操作维护人员最为基本的网元设 备操作维护需求, 但这种方式只是针对单独网元设备的检测, 而且是在网 元设备真正出现问题后才出现警示, 在用户日益关注设备自动化、 多网元 关联分析和故障预防的要求面前显然还是不够的, 在网络实际运行当中, 大部分问题的出现并非是所有网元设备都出现负载过重, 而有可能是由于 各个网元间的负载不均日益积累而造成的, 而网元负载不均在问题出现初 期可以通过网元间的指标差异分析进行发现并且可以给予及时的预防。
可以肯定地指出, 如果通过操作维护人员的手工来进行指标差异分析 显然是低效的, 且发现问题的实时性也很差, 并且随着网元管理系统所管 理的网元设备日益呈现复杂化和多元化, 网元设备之间的重要指标也越来 越多, 而且网元设备之间的关系也是越来越复杂, 从而使上述问题更为突 出, 由此网元管理系统在保证网元设备的可靠性、 可用性、 可维护性和稳 定性上正在接受着严峻的考验。 发明内容
为了解决现有技术中网元管理系统无法通过各网元设备指标自动化差 异分析来及时发现各网元设备是否存在安全隐患的问题, 本发明实施例的 目的在于提供一种网元负载不均检测处理方法、 装置及其系统。
为了达到本发明实施例的目的, 本发明实施例采用以下技术方案实现: 一种网元负载不均检测处理方法, 包括:
对各网元设备上报的网元性能指标根据预设的检测策略进行差异化分 析, 并将分析结果保存到网络指标緩冲池中;
定期从网络指标緩冲池中获取差异化分析结果, 并据此判断各网元设 备之间是否出现负载不均异常, 当判断得出出现异常时, 则发出警示消息。
优选地, 所述检测策略包括负载百分比门限策略、 约束绝对值门限策 略、 负载比例策略。 优选地, 所述警示消息包括当前指标信息、 门限信息、 负载比例信息。 优选地, 对各网元设备上报的网元性能指标根据预设的检测策略进行 差异化分析, 包括:
根据网元性能指标緩冲池中的指标信 ,包、判断网元管理系统所管理的各 网元设备的网元性能指标是否完备;
如果完备, 则判断各网元设备的网元性能指标是否都已超过预设的约 束绝对值门限;
如果是, 则根据各网元设备的网元性能指标和负载比例计算出负载均 值偏差, 将该负载均值偏差作为网元设备网元性能指标是否过负载的检测 基准, 其中, 负载均值偏差 P的计算公式如下:
P=∑Ci/∑Di, 其中, Ci为各网元性能指标的当前取值, Di为各网元设 备的负载比例;
根据各网元设备的网元性能指标和负载均值偏差计算出各个网元设备 的负载偏差百分比, 计算公式如下:
Vx=(Cx/Dx-P)*100/P, 其中, Vx为各网元设备的负载偏差百分比, Cx 为各网元性能指标的当前取值, Dx为各网元设备的负载比例, P为各网元 设备的负载均值偏差;
将各个网元设备的负载偏差百分比与预设的负载百分比门限进行比 较, 如果超过门限, 则标识相应的网元设备的网元性能指标为异常; 否贝 'J, 标识相应的网元设备的网元性能指标为正常。
优选地, 所述负载百分比门限包含正门限和负门限, 如果 Vx取值为正 值, 则将其与负载百分比门限的正门限进行比较, 如果 Vx取值为负值, 则 将其与负载百分比门限的负门限进行比较。
优选地, 当判断得出出现异常时, 发出警示消息, 包括:
检测先前是否已经发出警示并且警示还没有消除, 如果是不再发出冗 余的网元设备过载警示; 否则发出新的网元设备过载警示, 其中携带所述 当前指标信息、 门限信息和负载比例信息。
一种网元负载不均检测处理装置, 包括:
负载不均策略配置单元, 配置为对各网元设备网元性能指标进行差异 化分析的检测策略的参数配置;
网元性能指标緩冲池, 配置为存放各网元设备上报的网元性能指标; 负载不均检测单元, 配置为对各网元设备上报的网元性能指标根据预 设的检测策略进行差异化分析, 并将分析结果保存到网络指标緩冲池中; 网络指标緩冲池, 配置为存放各网元设备的网元性能指标在进行差异 化分析后的分析结果。
优选地, 所述检测策略包括负载百分比门限策略、 约束绝对值门限策 略、 负载比例策略。
一种网元负载不均检测处理系统, 包括:
网元设备, 配置为负责上报各自的网元性能指标;
网元负载不均检测处理装置, 配置为对各网元设备上报的网元性能指 标根据预设的检测策略进行差异化分析, 并将分析结果保存到网络指标緩 冲池中;
网元管理系统, 配置为定期从网元负载不均检测处理装置的网络指标 緩冲池中获取差异化分析结果, 并据此判断各网元设备之间是否出现负载 不均异常, 当判断得出出现异常时, 则发出警示消息。
优选地, 所述检测策略包括负载百分比门限策略、 约束绝对值门限策 略、 负载比例策略。
优选地, 所述警示消息包括当前指标信息、 门限信息、 负载比例信息。 优选地, 网元负载不均检测处理装置, 配置为根据网元性能指标緩冲 池中的指标信息判断网元管理系统所管理的各网元设备的网元性能指标是 否完备;
如果完备, 判断各网元设备的网元性能指标是否都已超过预设的约束 绝对值门限;
如果是, 则根据各网元设备的网元性能指标和负载比例计算出负载均 值偏差, 将该负载均值偏差作为网元设备网元性能指标是否过负载的检测 基准, 其中, 负载均值偏差 P的计算公式如下:
P=∑Ci/∑Di, 其中, Ci为各网元性能指标的当前取值, Di为各网元设 备的负载比例;
根据各网元设备的网元性能指标和负载均值偏差计算出各个网元设备 的负载偏差百分比, 计算公式如下:
Vx=(Cx/Dx-P)*100/P, 其中, Vx为各网元设备的负载偏差百分比, Cx 为各网元性能指标的当前取值, Dx为各网元设备的负载比例, P为各网元 设备的负载均值偏差;
将各个网元设备的负载偏差百分比与预设的负载百分比门限进行比 较, 如果超过门限, 则标识相应的网元设备的网元性能指标为异常; 否贝 'J, 标识相应的网元设备的网元性能指标为正常。
优选地, 所述负载百分比门限包含正门限和负门限, 如果 Vx取值为正 值, 则将其与负载百分比门限的正门限进行比较, 如果 Vx取值为负值, 则 将其与负载百分比门限的负门限进行比较。
优选地, 当网元管理系统, 配置为检测先前是否已经发出警示并且警 示还没有消除, 如果是则不再发出冗余的网元设备过载警示; 否则发出新 的网元设备过载警示, 其中携带所述当前指标信息、 门限信息和负载比例 信息。
通过上述本发明的技术方案可以看出, 本发明实施例采用负载均值偏 差计算、 负载偏差百分比计算以及与门限比对技术, 能够在多种网元设备 共管的场景下及时发现某些网元设备负载不均的问题, 从而通过发出警示 消息第一时间通知用户及时消除网元设备可能存在的安全隐患, 甚至采用 预设的自动化负载策略执行修改来自动消除隐患, 以提升网络通讯设备的 可靠性、 可维护性和稳定性。 附图说明
图 1为现有的网元负载过重处理构架示意图;
图 2 为本发明实施例提供的一种网元负载不均检测处理方法流程示意 图;
图 3为本发明实施例一种网元负载不均检测的组网示意图;
图 4为本发明实施例提供的一种网元负载不均检测处理系统结构示意 图;
图 5为本发明实施例对网元负载不均检测处理的方法流程图。 具体实施方式
下面结合附图和具体实施例对本发明所述技术方案作进一步的详细描 述。
如图 2所示, 本发明实施例提供的一种网元负载不均检测处理方法, 包括如下步骤:
S100、 对各网元设备上报的网元性能指标根据预设的检测策略进行差 异化分析, 并将分析结果保存到网络指标緩冲池中;
S200、 定期从网络指标緩冲池中获取差异化分析结果, 并据此判断各 网元设备之间是否出现负载不均异常, 当判断得出出现异常时, 则发出警 示消息。
具体地,在所述步骤 S100中,所述检测策略包括负载百分比门限策略、 约束绝对值门限策略、 负载比例策略。 具体地, 在所述步骤 S200中, 所述警示消息包括当前指标信息、 门限 信息、 负载比例信息。
具体地, 所述网元性能指标包含两种应用场景:
一种是检测同类网元的负载均衡, 例如检测 MSC ( Mobile Switching Center, 移动交换中心)池( POOL ) 中的各个 MSC网元的负载均衡, 检 测 SGSN ( Serving GPRS support Node, 服务 GPRS支持节点) POOL中的 各个 SGSN 网元的负载均衡, 所检测的指标可以是业务方面的指标, 例如 对 SGSN POOL检测的指标可以是附着用户数、 处于激活状态的会话平均 数、 Gn/Gp口流量等。
另一种是检测不同网元的负载均衡,例如检测 SGSN与 GGSN( Gateway
GPRS Support Node, 网关 GPRS支持节点)之间的负载均衡, 所检测的指 标一般是硬件相关的通用指标, 例如 CPU占用率、 内存占用率、 网络流量 等, 也可以是逻辑关联的业务指标, 例如 SGSN附着用户数和 GGSN承载 数之间的关联指标等, 其组网示意图如图 3 所示, 各网元提供对应指标取 值, 网元负载不均检测处理装置作为单独的设备或者作为网元管理系统中 的一个独立装置部署, 其可以对用户关心的部分网元进行负载均衡检测。
优选地, 在所述步骤 S100中, 对各网元设备上报的网元性能指标根据 预设的检测策略进行差异化分析的步骤包括:
A、根据网元性能指标緩冲池中的指标信息判断网元管理系统所管理的 各网元设备的网元性能指标是否完备, 如果完备, 则执行步骤 B, 否则执 行步骤 C;
B、判断各网元设备的网元性能指标是否都已超过预设的约束绝对值门 限, 如果是, 则执行步骤 D, 否则执行步骤 C;
C、 结束本次指标差异化分析;
D、 根据各网元设备的网元性能指标和负载比例计算出负载均值偏差, 将该负载均值偏差作为网元设备网元性能指标是否过负载的检测基准, 其 中, 负载均值偏差 P的计算公式如下:
P=∑Ci/∑Di, 其中, Ci为各网元性能指标的当前取值, Di为各网元设 备的负载比例;
E、根据各网元设备的网元性能指标和负载均值偏差计算出各个网元设 备的负载偏差百分比, 计算公式如下:
Vx=(Cx/Dx-P)*100/P, 其中, Vx为各网元设备的负载偏差百分比, Cx 为各网元性能指标的当前取值, Dx为各网元设备的负载比例, P为各网元 设备的负载均值偏差;
F、 将各个网元设备的负载偏差百分比与预设的负载百分比门限进行比 较, 如果超过门限, 则执行步骤 G, 否则执行步骤 H;
G、 标识相应的网元设备的网元性能指标为异常;
H、 标识相应的网元设备的网元性能指标为正常。
具体地, 在所述步骤 F 中, 负载百分比门限包含正门限和负门限, 如 果 Vx取值为正值, 则将其与负载百分比门限的正门限进行比较, 如果 Vx 取值为负值, 则将其与负载百分比门限的负门限进行比较。
具体地, 在所述步骤 S200中, 当判断得出出现异常时, 发出警示消息 的步骤包括:
a、 检测先前是否已经发出警示并且警示还没有消除, 如果是则执行步 骤 b, 否则执行步骤 c;
b、 不再发出冗余的网元设备过载警示;
c、 发出新的网元设备过载警示, 其中携带所述当前指标信息、 门限信 息和负载比例信息;
优选地, 所述方法还包括: 当判断得出正常时, 检测先前是否已经发 出警示并且警示还没有消除, 如果是则消除网元设备过载警示, 否则不做 任何处理。
具体地, 本发明实施例所提供的一种网元负载不均检测处理方法, 由 网元负载不均检测处理装置实现的该方法主要包括以下步骤:
步骤 1,根据设定的检测策略, 对各网元设备上报的网元性能指标进行 差异化分析, 并将差异化分析后的分析结果保存到网络指标緩冲池中, 检 测策略包括负载百分比门限、 约束绝对值门限、 负载比例。
本发明实施例中设定的检测策略包括以下三个部分:
第一部分是负载百分比门限, 包括正门限和负门限, 并且正门限和负 门限都可以配置多个级别门限, 配置为表明超过什么门限的负载不均指标 需要给出什么等级的警示;
第二部分是约束绝对值门限, 配置为表明多少基础之上的指标才需要 进行负载不均的检测, 因为在处理数据量不大的情况下各网元负载差异较 大是可以接受的;
第三部分是负载比例, 配置为表明各网元的处理能力如何, 因为各网 元由于其软 /硬件配置不同其处理能力也各不相同, 并非相同比例 1 :1 的关 系。
步骤 2,将网络指标緩冲池中的各网元设备差异化分析结果发送给网元 管理系统的网元管理者单元, 网元管理者单元发现结果异常, 向警示单元 发出警示消息, 警示信息中包括当前指标信息、 门限信息、 负载比例。
本发明实施例中警示信息包括以下三个部分:
第一部分是当前指标信息, 以便于操作维护人员实时了解当前的网元 设备具体负载情况;
第二部分是当前门限信息, 以便于操作维护人员了解警示消息发出时 的门限配置信息, 因为门限配置信息是可能会修改的, 所以相关信息需要 匹配; 第三部分是负载比例信息, 以便于操作维护人员了解警示消息发出时 的负载比例信息, 因为负载比例信息也是可以调整的。
本发明实施例还提供了一种网元负载不均检测处理装置, 参考图 4所 示, 其包括:
负载不均策略配置单元 32, 配置为对各网元设备网元性能指标进行差 异化分析的检测策略的参数配置;
网元性能指标緩冲池, 用于存放各网元设备上报的网元性能指标; 负载不均检测单元 31, 配置为对各网元设备上报的网元性能指标根据 预设的检测策略进行差异化分析, 并将分析结果保存到网络指标緩冲池中; 网络指标緩冲池 33, 配置为存放各网元设备的网元性能指标在进行差 异化分析后的分析结果。
例如, 在本发明实施例中, 网元性能指标緩冲池包括第一网元 M指标 緩冲池 34以及第一网元 N指标緩冲池 35, 其中:
第一网元 M指标緩冲池 34, 配置为存放网元 M的性能指标, 作为各 网元设备性能指标进行差异化分析的素材;
第一网元 N指标緩冲池 35, 配置为存放网元 N的性能指标, 作为各网 元设备性能指标进行差异化分析的素材;
所述检测策略包括负载百分比门限策略、 约束绝对值门限策略、 负载 比例策略。
在实际应用中,所述负载不均检测单元 31、负载不均策略配置单元 32、 网络指标緩冲池 33、 第一网元 M指标緩冲池 34和第一网元 N指标緩冲池 35 可以由中央处理器 (CPU, Central Processing Unit )、 数字信号处理器 ( DSP , Digital Signal Processor ) 或可编程逻辑阵列 (FPGA, Field - Programmable Gate Array ) 实现。
参考图 4所示, 本发明实施例提供了一种网元负载不均检测处理系统, 包括:
网元设备 10, 配置为负责上报各自的网元性能指标;
网元负载不均检测处理装置 30, 配置为对各网元设备上报的网元性能 指标根据预设的检测策略进行差异化分析, 并将分析结果保存到网络指标 緩冲池中;
网元管理系统 20,配置为定期从网元负载不均检测处理装置 30的网络 指标緩冲池中获取差异化分析结果, 并据此判断各网元设备之间是否出现 负载不均异常, 当判断得出出现异常时, 则发出警示消息。
具体地, 所述检测策略包括负载百分比门限策略、 约束绝对值门限策 略、 负载比例策略。
具体地, 所述警示消息包括当前指标信息、 门限信息、 负载比例信息。 具体地, 网元负载不均检测处理装置 30, 配置为根据网元性能指标緩 冲池中的指标信息判断网元管理系统所管理的各网元设备的网元性能指标 是否完备;
如果完备, 则判断各网元设备的网元性能指标是否都已超过预设的约 束绝对值门限;
如果是, 根据各网元设备的网元性能指标和负载比例计算出负载均值 偏差, 将该负载均值偏差作为网元设备网元性能指标是否过负载的检测基 准, 其中, 负载均值偏差 P的计算公式如下:
P=∑Ci/∑Di, 其中, Ci为各网元性能指标的当前取值, Di为各网元设 备的负载比例;
E、根据各网元设备的网元性能指标和负载均值偏差计算出各个网元设 备的负载偏差百分比, 计算公式如下:
Vx=(Cx/Dx-P)*100/P, 其中, Vx为各网元设备的负载偏差百分比, Cx 为各网元性能指标的当前取值, Dx为各网元设备的负载比例, P为各网元 设备的负载均值偏差;
将各个网元设备的负载偏差百分比与预设的负载百分比门限进行比 较, 如果超过门限, 则执行步骤 G, 否则执行步骤 H;
标识相应的网元设备的网元性能指标为异常;
标识相应的网元设备的网元性能指标为正常。
所述负载百分比门限包含正门限和负门限, 如果 Vx取值为正值, 则将 其与负载百分比门限的正门限进行比较, 如果 Vx取值为负值, 则将其与负 载百分比门限的负门限进行比较。
具体地, 网元管理系统 20配置为检测先前是否已经发出警示并且警示 还没有消除, 如果是不再发出冗余的网元设备过载警示; 否则发出新的网 元设备过载警示, 其中携带所述当前指标信息、 门限信息和负载比例信息。
具体地, 本实施例中, 用以实现图 2 所示的网元负载均衡检测处理方 法的网元负载不均检测处理装置 30, 如图 4所示, 该装置包括: 相互连接 的负载不均检测单元 31和负载不均策略配置单元 32、相互连接的负载不均 检测单元 31和网络指标緩冲池 33、 相互连接的网络指标緩冲池 33和第一 网元 M指标緩冲池 34和第一网元 N指标緩冲池 35。
其中负载不均检测单元 31, 配置为根据配置的检测策略, 对网元设备 上报的网元设备网元性能指标进行差异化负载均衡分析, 并将差异化分析 结果保存到网络指标緩冲池 33中。
负载不均策略配置单元 32, 配置为配置和存储负载不均的检测策略参 数。
网络指标緩冲池 33, 配置为保存经过差异化分析的各网元设备负载均 衡检测结果, 并且将差异化分析结果发送给网元管理系统 20。
第一网元 M指标緩冲池 34, 配置为保存经过差异化分析的 M网元设 备负载均衡检测结果。 第一网元 N指标緩冲池 35,配置为保存经过差异化分析的 N网元设备 负载均衡检测结果。
与负载不均检测单元 31相连的网元设备 10包括: 处理单元 11、 网元 代理者单元 12和网元设备指标緩冲池 13。
其中, 处理单元 11, 配置为检测网元设备的运行状态, 当网络上发生 业务时向网元代理者单元 12上报对应的数据信息。
网元代理者 12, 与处理单元 11和负载不均检测单元 31相连, 配置为 接收处理单元 11上报的数据信息并提取成网元性能指标后緩存到网元设备 的网元设备指标緩冲池 13中, 并且适当的时候(定时)把网元设备的网元 性能指标上报给负载不均检测单元 31进行性能指标差异化分析, 原始的网 元设备的网元性能指标也会上报一份给网元管理系统 20, 用于单独网元的 网元设备指标分析。
网元设备指标緩冲池 13, 与网元代理者单元 12相连, 配置为对网元设 备的性能指标进行緩存。
与网络指标緩冲池 33相连的网元管理系统 20包括: 网元管理者单元 21,第二网元 M指标緩冲池 22、第二网元 N指标緩冲池 23和警示单元 24。
其中, 网元管理者单元 21, 与网元代理者 12和网络指标緩冲池 33相 连, 配置为接收网元代理者 12发送过来的单独网元的网元设备的网元性能 指标, 緩存到第二网元 M指标緩冲池和第二网元 N指标緩冲池中, 并对其 进行分析处理; 以及进一步配置为从网络指标緩冲池 33中获取各网元设备 的差异化负载均衡检测结果并进行分析处理, 如果发现异常则向警示单元 24发出警示。
第二网元 M指标緩冲池 22, 与网元管理者单元 21相连, 配置为对网 元 M的网元设备指标进行緩存。
第二网元 N指标緩冲池 23, 与网元管理者单元 21相连, 配置为对网 元 N的网元设备指标进行緩存。
警示单元 24, 与网元管理者单元 21 相连, 配置为对网元管理者单元 21发送过来的警示消息进行显示和声光等警示。
在实际应用中, 所述网元管理者单元 21, 第二网元 M指标緩冲池 22、 第二网元 N指标緩冲池 23和警示单元 24可以由 CPU、DSP或 FPGA实现。
下面结合图 4所示的网元负载均衡检测装置, 对本发明实施例中网元 负载不均检测处理方法分别进行详细阐述。 如图 5所示, 图 5为本发明实 施例对网元负载不均检测处理的方法流程图, 该方法主要包括以下步骤: 其中, 检测过程包括:
步骤 401,网元代理者单元 12向负载不均检测单元 31上报其网元的网 元设备性能指标, 其指标中必须包含有效的性能指标取值。
步骤 402, 负载不均检测单元 31根据策略检测上报的网元设备指标是 否完备, 即是否策略监控的网元都已上报性能指标, 如果是, 则执行步骤 403, 如果否, 则流程结束。
步骤 403, 负载不均检测单元 31根据策略检测上报的各网元设备指标 是否超过策略要求的约束绝对值门限, 如果是, 则执行步骤 404, 如果否, 则流程结束。
步骤 404, 负载不均检测单元 31根据各网元设备指标取值和策略配置 的各网元设备负载比例计算出负载均值偏差? =∑(^ /∑01, 其中 Ci为各网 元检测指标的当前取值, Di为各网元的负载比例。
步骤 405, 负载不均检测单元 31检查各网元设备指标是否都已经检测 完毕, 如果是, 则流程结束, 如果否, 则执行步骤 406。
步骤 406, 负载不均检测单元 31取出一个未检测的网元设备指标。 步骤 407,负载不均检测单元 31计算负载偏差百分比 V = (C/D - P)*100 / P, 其中 C为此网元检测指标的当前取值, D为此网元的负载比例, P为 负载均值偏差。
步骤 408, 负载不均检测单元 31检测此网元的负载偏差百分比是否超 过策略配置的负载百分比门限, 如果是, 则执行步骤 409, 如果否, 则执行 步骤 411。
步骤 409, 负载不均检测单元 31把该网元的网元设备指标标识为异常 并放入网络指标緩冲池 33中, 重复执行步骤 405。
步骤 410,负载不均检测单元 31对网络指标緩冲池 33中的网元设备指 标进行老化处理。
步骤 411, 负载不均检测单元 31把该网元的网元设备指标标识为正常 并放入网络指标緩冲池 33中, 重复执行步骤 405。
其中, 处理过程包括:
步骤 412,网元管理者单元 21从网络指标緩冲池 33中提取网元设备的 差异化负载均衡检测结果并进行分析处理,如果发现异常,则进入步骤 413; 如果发现正常, 则进入步骤 414;
步骤 413、网元管理者单元 21从网络指标緩冲池 33中提取网元设备指 标, 检测是否已经发出过对应的警示并且该警示还没有消除, 如果是, 则 流程结束, 如果否, 则执行步骤 415。
步骤 414,网元管理者单元 21从网络指标緩冲池 33中提取网元设备指 标, 检测是否已经发出过对应的警示并且该警示还没有消除, 如果是, 则 执行步骤 416, 如果否, 则流程结束。
步骤 415,网元管理者单元 21向警示单元 24发出新的网元设备负载不 均警示消息, 警示消息中携带当前指标信息、 门限信息、 负载比例信息。
步骤 416,网元管理者单元 21向警示单元 24发出消除网元设备负载不 均警示的警示去除消息。
需要指出的是, 本发明实施例还能解决其他类似的问题, 例如, 将网 元内的类似指标进行负载不均的检测来自动发掘网元内是否存在逻辑单元 负载不均的问题等。
另外本发明还有其他多种实施例, 例如, 将负载不均检测单元不通过 网元代理者单元, 直接与网元管理者单元相连等等。 总之在未背离本发明 精神及其实质的情况下, 熟悉本领域的技术人员当可根据本发明做出各种 相应的改变和变形, 但这些相应的改变和变形都应属于本发明的保护范围。
以上所述仅为本发明的优选实施例, 并非因此限制本发明的专利范围, 凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换, 或直 接或间接运用在其他相关的技术领域, 均同理包括在本发明的专利保护范 围内。

Claims

权利要求书
1、 一种网元负载不均检测处理方法, 所述方法包括:
对各网元设备上报的网元性能指标根据预设的检测策略进行差异化分 析, 并将分析结果保存到网络指标緩冲池中;
定期从网络指标緩冲池中获取差异化分析结果, 并据此判断各网元设 备之间是否出现负载不均异常, 当判断得出出现异常时, 则发出警示消息。
2、 如权利要求 1所述的网元负载不均检测处理方法, 其中, 所述检测 策略包括负载百分比门限策略、 约束绝对值门限策略、 负载比例策略。
3、 如权利要求 1所述的网元负载不均的检测处理方法, 其中, 所述警 示消息包括当前指标信息、 门限信息、 负载比例信息。
4、 如权利要求 2所述的网元负载不均检测处理方法, 其中, 对各网元 设备上报的网元性能指标根据预设的检测策略进行差异化分析, 包括: 根据网元性能指标緩冲池中的指标信 ,包、判断网元管理系统所管理的各 网元设备的网元性能指标是否完备;
如果完备, 则判断各网元设备的网元性能指标是否都已超过预设的约 束绝对值门限;
如果都已超过预设的约束绝对值门限, 则根据各网元设备的网元性能 指标和负载比例计算出负载均值偏差, 将该负载均值偏差作为网元设备网 元性能指标是否过负载的检测基准, 其中, 负载均值偏差 P的计算公式如 下:
P=∑Ci/∑Di, 其中, Ci为各网元性能指标的当前取值, Di为各网元设 备的负载比例;
根据各网元设备的网元性能指标和负载均值偏差计算出各个网元设备 的负载偏差百分比, 计算公式如下:
Vx=(Cx/Dx-P)*100/P, 其中, Vx为各网元设备的负载偏差百分比, Cx 为各网元性能指标的当前取值, Dx为各网元设备的负载比例, P为各网元 设备的负载均值偏差;
将各个网元设备的负载偏差百分比与预设的负载百分比门限进行比 较, 如果超过门限, 则标识相应的网元设备的网元性能指标为异常; 如果 不超过门限, 则标识相应的网元设备的网元性能指标为正常。
5、 如权利要求 4所述的网元负载不均检测处理方法, 其中, 所述负载 百分比门限包含正门限和负门限, 如果 Vx取值为正值, 则将其与负载百分 比门限的正门限进行比较, 如果 Vx取值为负值, 则将其与负载百分比门限 的负门限进行比较。
6、 如权利要求 1所述的网元负载不均检测处理方法, 其中, 当判断得 出出现异常时, 发出警示消息, 包括:
检测先前是否已经发出警示并且警示还没有消除, 如果是, 则不再发 出冗余的网元设备过载警示; 如果否, 则发出新的网元设备过载警示。
7、 一种网元负载不均检测处理装置, 所述装置包括:
负载不均策略配置单元, 配置为对各网元设备网元性能指标进行差异 化分析的检测策略的参数配置;
网元性能指标緩冲池, 配置为存放各网元设备上报的网元性能指标; 负载不均检测单元, 配置为对各网元设备上报的网元性能指标根据预 设的检测策略进行差异化分析, 并将分析结果保存到网络指标緩冲池中; 网络指标緩冲池, 配置为存放各网元设备的网元性能指标在进行差异 化分析后的分析结果。
8、 如权利要求 7所述的网元负载不均检测处理装置, 其中, 所述检测 策略包括负载百分比门限策略、 约束绝对值门限策略、 负载比例策略。
9、 一种网元负载不均检测处理系统, 所述系统包括:
网元设备, 配置为负责上报各自的网元性能指标; 网元负载不均检测处理装置, 配置为对各网元设备上报的网元性能指 标根据预设的检测策略进行差异化分析, 并将分析结果保存到网络指标緩 冲池中;
网元管理系统, 配置为定期从网元负载不均检测处理装置的网络指标 緩冲池中获取差异化分析结果, 并据此判断各网元设备之间是否出现负载 不均异常, 当判断得出出现异常时, 则发出警示消息。
10、 如权利要求 9所述的网元负载不均检测处理系统, 其中, 所述检 测策略包括负载百分比门限策略、 约束绝对值门限策略、 负载比例策略。
11、 如权利要求 9 所述的网元负载不均检测处理系统, 其中, 所述警 示消息包括当前指标信息、 门限信息、 负载比例信息。
12、 如权利要求 10所述的网元负载不均检测处理系统, 其中, 网元负 载不均检测处理装置, 配置为根据网元性能指标緩冲池中的指标信息判断 网元管理系统所管理的各网元设备的网元性能指标是否完备;
如果完备, 则判断各网元设备的网元性能指标是否都已超过预设的约 束绝对值门限;
如果是, 则根据各网元设备的网元性能指标和负载比例计算出负载均 值偏差, 将该负载均值偏差作为网元设备网元性能指标是否过负载的检测 基准, 其中, 负载均值偏差 P的计算公式如下:
P=∑Ci/∑Di, 其中, Ci为各网元性能指标的当前取值, Di为各网元设 备的负载比例;
E、根据各网元设备的网元性能指标和负载均值偏差计算出各个网元设 备的负载偏差百分比, 计算公式如下:
Vx=(Cx/Dx-P)*100/P, 其中, Vx为各网元设备的负载偏差百分比, Cx 为各网元性能指标的当前取值, Dx为各网元设备的负载比例, P为各网元 设备的负载均值偏差; 将各个网元设备的负载偏差百分比与预设的负载百分比门限进行比 较, 如果超过门限, 则标识相应的网元设备的网元性能指标为异常; 如果 不超过门限, 标识相应的网元设备的网元性能指标为正常。
13、 如权利要求 12所述的网元负载不均检测处理系统, 其中, 所述负 载百分比门限包含正门限和负门限, 如果 VX取值为正值, 则将其与负载百 分比门限的正门限进行比较, 如果 Vx取值为负值, 则将其与负载百分比门 限的负门限进行比较。
14、 如权利要求 9所述的网元负载不均检测处理系统, 其中, 所述网 元管理系统, 配置为检测先前是否已经发出警示并且警示还没有消除, 如 果是, 则不再发出冗余的网元设备过载警示; 否则, 发出新的网元设备过 载警示。
PCT/CN2013/084512 2012-10-22 2013-09-27 网元负载不均检测处理方法、装置及其系统 WO2014063557A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP13849007.3A EP2899918A4 (en) 2012-10-22 2013-09-27 METHOD, DEVICE AND SYSTEM FOR DETECTING THE LOADING WEIGHT OF A NETWORK ELEMENT
US14/437,337 US9674065B2 (en) 2012-10-22 2013-09-27 Method, apparatus and system for detecting network element load imbalance

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201210404787.9A CN102932170B (zh) 2012-10-22 2012-10-22 网元负载不均检测处理方法、装置及其系统
CN201210404787.9 2012-10-22

Publications (1)

Publication Number Publication Date
WO2014063557A1 true WO2014063557A1 (zh) 2014-05-01

Family

ID=47646877

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/084512 WO2014063557A1 (zh) 2012-10-22 2013-09-27 网元负载不均检测处理方法、装置及其系统

Country Status (4)

Country Link
US (1) US9674065B2 (zh)
EP (1) EP2899918A4 (zh)
CN (1) CN102932170B (zh)
WO (1) WO2014063557A1 (zh)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102932170B (zh) * 2012-10-22 2016-06-22 中兴通讯股份有限公司 网元负载不均检测处理方法、装置及其系统
CN105323100B (zh) * 2014-07-31 2019-10-11 南京中兴软件有限责任公司 告警的生成方法及装置
CN108259255B (zh) * 2016-12-29 2020-05-15 大唐移动通信设备有限公司 一种网元数据分析方法及装置
US11157323B2 (en) * 2019-01-10 2021-10-26 International Business Machines Corporation Multiple metric based load prediction and resource allocation in an active stream processing job
US11630927B2 (en) * 2019-10-15 2023-04-18 Ruentex Engineering & Construction Co., Ltd. Method and system for processing building energy information
CN113406439B (zh) * 2021-06-23 2024-08-23 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 一种配电网故障定位方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101384054A (zh) * 2007-09-04 2009-03-11 中兴通讯股份有限公司 一种通过性能数据监测网络异常情况的方法
CN101458634A (zh) * 2008-01-22 2009-06-17 中兴通讯股份有限公司 负载均衡调度方法和装置
CN101945407A (zh) * 2010-10-22 2011-01-12 东南大学 一种应用于移动业务内容监控的负载均衡方法
CN102232282A (zh) * 2010-10-29 2011-11-02 华为技术有限公司 一种实现数据中心资源负载均衡的方法及装置
CN102932170A (zh) * 2012-10-22 2013-02-13 中兴通讯股份有限公司 网元负载不均检测处理方法、装置及其系统

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FI107312B (fi) * 1997-10-14 2001-06-29 Nokia Networks Oy Verkonvalvontamenetelmä tietoliikenneverkkoa varten
DE60144446D1 (de) 2001-08-24 2011-05-26 Ericsson Telefon Ab L M Verfahren und mittel zur weiterverteilung von teilnehmerinformationen in umts-netzwerken, in denen die knoten in pools angeordnet sind
US8068408B2 (en) * 2004-11-01 2011-11-29 Alcatel Lucent Softrouter protocol disaggregation
DE102005056250A1 (de) * 2005-11-25 2007-05-31 Siemens Ag Verfahren bzw. Computerprogrammprodukt zur Bestimmung der Leistung eines Computersystems
CN101136805B (zh) * 2007-05-30 2010-06-09 中兴通讯股份有限公司 一种性能告警系统及其性能门限的获取方法
US9274842B2 (en) * 2010-06-29 2016-03-01 Microsoft Technology Licensing, Llc Flexible and safe monitoring of computers
CN102611564B (zh) * 2011-01-25 2017-04-05 中兴通讯股份有限公司 一种实现性能告警的方法、系统及网元
US9125173B2 (en) * 2012-02-29 2015-09-01 Via Telecom Co., Ltd. Device and method for access probe enhancements
US20140105218A1 (en) * 2012-10-12 2014-04-17 Prashant H. Anand Queue monitoring to filter the trend for enhanced buffer management and dynamic queue threshold in 4g ip network/equipment for better traffic performance

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101384054A (zh) * 2007-09-04 2009-03-11 中兴通讯股份有限公司 一种通过性能数据监测网络异常情况的方法
CN101458634A (zh) * 2008-01-22 2009-06-17 中兴通讯股份有限公司 负载均衡调度方法和装置
CN101945407A (zh) * 2010-10-22 2011-01-12 东南大学 一种应用于移动业务内容监控的负载均衡方法
CN102232282A (zh) * 2010-10-29 2011-11-02 华为技术有限公司 一种实现数据中心资源负载均衡的方法及装置
CN102932170A (zh) * 2012-10-22 2013-02-13 中兴通讯股份有限公司 网元负载不均检测处理方法、装置及其系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP2899918A4 *

Also Published As

Publication number Publication date
CN102932170A (zh) 2013-02-13
EP2899918A1 (en) 2015-07-29
CN102932170B (zh) 2016-06-22
US9674065B2 (en) 2017-06-06
US20150271045A1 (en) 2015-09-24
EP2899918A4 (en) 2015-10-21

Similar Documents

Publication Publication Date Title
WO2014063557A1 (zh) 网元负载不均检测处理方法、装置及其系统
WO2019169743A1 (zh) 一种服务器故障的检测方法及系统
KR100561628B1 (ko) 통계적 분석을 이용한 네트워크 수준에서의 이상 트래픽감지 방법
CN111092786B (zh) 网络设备安全认证服务可靠性增强系统
US11050609B2 (en) Technique for reporting and processing alarm conditions occurring in a communication network
CN105791033A (zh) 一种调控服务器运行状态的方法、装置及系统
US20110185235A1 (en) Apparatus and method for abnormality detection
EP2807563B1 (en) Network debugging
CN107547228B (zh) 一种基于大数据的安全运维管理平台的实现架构
CN107465575A (zh) 一种集群的监控方法及系统
CN106487612A (zh) 一种服务器节点监控方法、监控服务器及系统
TW200637242A (en) Method and apparatus for determining impact of faults on network service
CN113612647B (zh) 一种告警处理方法及装置
CN110730087A (zh) 一种告警风暴的处理方法及装置
WO2018035765A1 (zh) 网络异常的检测方法及装置
CN102104606B (zh) 一种内网蠕虫主机检测方法
CN105897492A (zh) 一种云数据中心监控系统
CN103457792B (zh) 一种故障检测方法和装置
KR20050090640A (ko) 유해 트래픽 분석 시스템 및 방법
CN111277805B (zh) 一种针对olt上联链路crc错误的处理方法及装置
WO2015131708A1 (zh) 一种监控终端设备的方法、装置及系统
JP2011114822A (ja) ネットワーク管理装置およびネットワーク管理方法
CN112001623B (zh) 软件负载均衡的健康度的评估方法、系统、介质和设备
TW201513690A (zh) 遞迴式異常網路流量偵測方法
KR101472737B1 (ko) 인라인 장비의 장애 감지를 통한 패킷 바이패스 장치 및 그 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13849007

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 14437337

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2013849007

Country of ref document: EP