WO2012031439A1 - 网管系统性能数据汇总的方法和装置 - Google Patents

网管系统性能数据汇总的方法和装置 Download PDF

Info

Publication number
WO2012031439A1
WO2012031439A1 PCT/CN2010/080112 CN2010080112W WO2012031439A1 WO 2012031439 A1 WO2012031439 A1 WO 2012031439A1 CN 2010080112 W CN2010080112 W CN 2010080112W WO 2012031439 A1 WO2012031439 A1 WO 2012031439A1
Authority
WO
WIPO (PCT)
Prior art keywords
performance data
message queue
thread
data summary
management system
Prior art date
Application number
PCT/CN2010/080112
Other languages
English (en)
French (fr)
Inventor
熊纪涛
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
Publication of WO2012031439A1 publication Critical patent/WO2012031439A1/zh

Links

Classifications

    • 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

Definitions

  • the present invention relates to the field of communications, and in particular to a method and apparatus for summarizing performance data of a network management system. Background technique
  • Performance management is one of the important functions in telecommunication network management.
  • the purpose of performance management is to perform performance monitoring on networks, network elements or devices, collect relevant performance statistics, evaluate the effectiveness of networks and network elements, and report on telecommunication equipment. Status, as well as support for network planning and network analysis.
  • the network management system needs to summarize the performance data into data of different time granularity, such as hourly, day, week, month, etc., so that there is direct data extraction during query, and the efficiency of the network management system is improved. .
  • the performance data summary method usually uses parallel summarization.
  • the traditional parallel aggregation method is to collect a batch of threads to collect the performance data after the network management system collects a batch of performance data.
  • the above parallel summary method has the following drawbacks: Due to the large scale, high concurrency, and high real-time characteristics of the telecommunication network, when a large amount of performance data arrives, the network management system needs to establish a large number of threads for aggregation, which is easy to generate parallel storms. The network management system crashed. Summary of the invention
  • the main object of the present invention is to provide a method and apparatus for summarizing performance data of a network management system, by preventing a message queue including a performance data summary task, and limiting the number of threads executing a performance data summary task, thereby preventing concurrent threads due to a large number of threads.
  • the resulting network management system crashed.
  • a method for summarizing performance data of a network management system comprising: establishing a plurality of message queues, The message queue includes a performance data summary task; establishing a thread whose number does not exceed a preset value, the thread acquiring a message queue, and performing a performance data summary task in the message queue.
  • the establishing a plurality of message queues includes: classifying the message queue according to the category of the performance data summary task; adding the newly added performance data summary task to the corresponding message queue according to the category.
  • the added performance data summary task is added to the corresponding message queue according to the category, including: determining a new performance data summary task category;
  • the method further includes: periodically querying whether the message queue is full; adjusting the number of threads and/or waiting according to whether there is a full message queue Time, the waiting time is the time interval at which the thread acquires the message queue.
  • the adjusting the number of threads and/or the waiting time according to whether there is a full message queue includes:
  • a device for summarizing performance data of a network management system comprising: Establishing a queue module, configured to establish multiple message queues, where the message queue includes a performance data summary task;
  • a thread module is configured to establish a thread whose number does not exceed a preset value, and the thread acquires a message queue, and is used to perform a performance data summary task in the message queue.
  • the establishing queue module includes:
  • a classification unit configured to classify the message queue by the category of the performance data summary task; and adding a task unit for adding the new performance data summary task to the corresponding message queue according to the category.
  • the adding task unit includes:
  • a judging subunit configured to determine a category of the newly added performance data summary task
  • a first adding sub-unit configured to add a new performance data summary task to the message queue when the message queue of the category to which the newly added performance data summary task belongs is not satisfied
  • the second adding sub-unit is configured to stop the enqueuing when the message queue of the category to which the new performance data summary task belongs is full, wait for the message queue to be taken away, create a new message queue, and add a new performance data summary task. To the newly created message queue;
  • the third adding sub-unit is configured to create a new message queue when the message queue of the category to which the new performance data summary task belongs does not exist, and add a new performance data summary task to the newly created message queue.
  • the device further includes:
  • a query module configured to periodically query whether the message queue is full
  • the adjusting thread module is configured to adjust the number of threads and/or the waiting time according to whether there is a full message queue, where the waiting time is a time interval at which the thread acquires the message queue.
  • the adjustment thread module includes:
  • a first adjusting unit configured to reduce a waiting time of the thread when there is a full message queue, and trigger the establishing thread mode when a waiting time of the thread reaches a preset minimum value a new thread is created in the block until the number of threads reaches the preset value;
  • the second adjusting unit is configured to increase the waiting time of the thread when there is no full message queue, and end the thread when the waiting time of the thread reaches a preset maximum value.
  • the method and device for summarizing performance data of a network management system provided by the present invention first establish a plurality of message queues including performance data summary tasks, and then establish a thread execution performance data summary task that does not exceed a preset value. It limits the number of threads working in parallel, preventing the network management system from crashing due to a large number of thread concurrency.
  • FIG. 1 is a schematic flowchart of a method for summarizing performance data of a network management system according to an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of establishing a message queue in an embodiment of a method for summarizing performance data of a network management system according to the present invention
  • FIG. 3 is a schematic flowchart of adding a new performance data summary task to a message queue in a method for summarizing performance data of a network management system according to an embodiment of the present invention
  • FIG. 4 is a schematic flow chart of still another embodiment of a method for summarizing performance data of a network management system according to the present invention.
  • FIG. 5 is a schematic flowchart of adjusting a thread in another embodiment of a method for summarizing performance data of a network management system according to the present invention
  • FIG. 6 is a schematic structural diagram of an apparatus for synthesizing performance data of a network management system according to the present invention
  • FIG. 7 is a schematic structural diagram of establishing a queue module in an embodiment of a device for collecting performance data of a network management system according to the present invention
  • FIG. 8 is a schematic structural diagram of adding a task unit in an embodiment of a device for collecting performance data of a network management system according to the present invention.
  • FIG. 9 is a schematic structural diagram of still another embodiment of a device for collecting performance data of a network management system according to the present invention.
  • FIG. 10 is a schematic diagram of adjusting a thread in another embodiment of a device for collecting performance data of a network management system according to the present invention. Schematic diagram of the module. detailed description
  • the method and device for summarizing the performance data of the network management system provided by the present invention firstly establish a plurality of message queues including performance data summary tasks, and then establish a thread execution performance data summary task that does not exceed a preset value, thereby limiting parallel work.
  • a method for summarizing performance data of a network management system includes: Step 101: Establish a plurality of message queues, where the message queue includes a performance data summary task.
  • the network management system first establishes multiple message queues.
  • the message queue includes performance data summary tasks.
  • Each performance data summary task includes information about performance data collected by the network management system, such as the time and location of performance data.
  • Message queues are non-linear, and the length of each message queue can be the same or different, and can be preset and adjusted as needed to accommodate a certain number of performance data aggregation tasks.
  • Step 102 Establish a thread whose number does not exceed a preset value, the thread acquires a message queue, and performs a performance data summary task in the message queue.
  • the network management system then establishes a thread that performs a performance data summary task.
  • the maximum number of threads is a preset value to prevent the network management system from collapsing when a large number of threads are concurrent.
  • each thread works independently.
  • Each thread can acquire one or more pending message queues ⁇ ij periodically or randomly.
  • one thread processes one message queue to be processed at a time, and then executes the message queue in turn.
  • the last thread can destroy the message queue and continue to acquire the next pending message queue.
  • the performance data summary task performed by the thread is mainly to aggregate and analyze multiple performance data of the same granularity such as one hour or one day, and complete functions such as summation, averaging, maximization, and minimum.
  • each thread can process one or more message queues at a time according to the capability, preventing each thread from being overburdened.
  • the network management system can be prevented from collapsing when a large number of threads are concurrent.
  • the number of threads working in parallel is preset in advance by the network management system, and can be adjusted according to the capabilities of the network management system.
  • the method for summarizing the performance data of the network management system solves the network management system crash caused by a large number of thread concurrency by establishing a message queue including a performance data summary task and limiting the number of threads for performing parallel work of the performance data summary task. .
  • step 101 may include:
  • Step 1011 Sort the message queue according to the category of the performance data summary task.
  • the network management system can classify the message queue according to certain characteristics of the performance data.
  • the measurement type of the performance data is taken as an example to classify the message queue.
  • step 1012 the newly added performance data summary task is added to the corresponding message queue according to the category.
  • the new performance data aggregation task is inserted into the tail of the corresponding message queue according to the measurement type, waiting for the thread to execute.
  • step 1012 may include:
  • Step 1021 determining a new performance data summary task category
  • Step 1022 When the message queue of the category to which the new performance data summary task belongs is not full, add a new performance data summary task to the message queue; or,
  • Step 1023 When the message queue of the category to which the new performance data summary task belongs is full, stop enqueuing, wait for the message queue to be taken away, create a new message queue, and add a new performance data summary task to the newly created message queue. ; or,
  • Step 1024 When the message queue of the category to which the new performance data summary task belongs does not exist, create a new message queue, and add a new performance data summary task to the newly created message queue.
  • the network management system determines the category of the new performance data summary task, such as the measurement type CS Basic.
  • the performance data summary task of the measurement arrives, if the corresponding message queue is not full, the performance data summary task is inserted into the tail of the corresponding message queue, and the thread is executed.
  • the new performance data summary task is stopped, and after waiting for the thread to take out the corresponding message queue, a new message queue of measurement type CS Basic Measurement is created and will wait. And the subsequent arrival measurement type is the performance data summary task of CS Basic Measurement added to the newly created message queue.
  • step 102 after performing step 102, the method further includes:
  • Step 103 Timely query whether the message queue is full.
  • the network management system can periodically query whether all message queues are full.
  • the network management system can implement the query function by setting a counter: When the message queue is full, the counter is incremented by 1. When the thread takes out a full message queue, the counter is decremented by one, and the network management system can periodically obtain the counter count, such as The counter count is non-zero, indicating that at least one message queue is full. When the counter is 0, it means that all message queues are not full.
  • the network management system can also query by setting the full identifier for each message queue, or by other means to achieve the full status of the query message queue.
  • Step 104 Adjust the number of threads and/or waiting time according to whether there is a full message queue, and the waiting time is a time interval for the thread to acquire the message queue.
  • the network management system can reduce the waiting time of the thread and/or increase the number of threads working in parallel to speed up the execution speed of the performance data summary task.
  • the waiting time is the time interval at which the thread gets the message queue.
  • step 104 may include:
  • Step 1041 When there is a full message queue, reduce the waiting time of the thread. When the waiting time of the thread reaches a preset minimum value, create a new thread until the number of threads reaches a preset value.
  • Step 1042 When there is no full message queue, increase the waiting time of the thread, and when the waiting time of the thread reaches the preset maximum value, the thread ends.
  • the network management system can adjust the number of threads and the waiting time individually or simultaneously, and other ways that can be implemented by those skilled in the art to adjust the number of threads according to the performance data aggregation task can be applied to the present invention.
  • the method for summarizing the performance data of the network management system provided by the present invention prevents the network management system from crashing due to a large number of thread concurrency by establishing a message queue including the performance data summary task and limiting the number of threads performing the performance data summary task. Further, the thread can be adjusted according to the number of performance data summary tasks to improve the performance of the network management system.
  • an apparatus for summarizing the performance data of the network management system of the present invention includes: establishing a queue module 10 for establishing a plurality of message queues, wherein the message queue includes a performance data aggregation task;
  • the thread module 20 is configured to establish a thread whose number does not exceed a preset value, and the thread acquires a message queue for performing performance data summary tasks in the message queue.
  • the queue module 10 is first established with multiple message queues.
  • the message queue includes performance data summary tasks.
  • Each performance data summary task includes information about performance data collected by the network management system, such as the time and location of performance data.
  • Message queues are non-linear. Each message queue can be the same or different in length. It can be pre-set and adjusted as needed to accommodate a certain number of performance data aggregation tasks.
  • the threading module 20 then establishes a thread that performs a performance data summary task.
  • the maximum number of threads is a preset value to prevent the network management system from crashing when a large number of threads are concurrent.
  • each thread works independently.
  • Each thread can acquire one or more message queues to be processed periodically or randomly.
  • one thread processes one message queue to be processed at a time, and then sequentially executes the message queues.
  • the performance data summary task until all the performance data summary tasks in the message queue are executed, the last thread can destroy the message queue and continue to acquire the next pending message queue.
  • the performance data summary task performed by the thread is mainly to aggregate and analyze multiple performance data of the same granularity, such as one hour or one day, to complete the functions of summation, averaging, maximization, and minimum.
  • each thread can process one or more message queues at a time according to the capability, preventing each thread from being overburdened, and preventing a large number of threads from working in parallel as a preset value.
  • the network management system crashed when threads were concurrency.
  • the number of threads working in parallel is preset in advance by the network management system, and can be adjusted according to the capabilities of the network management system.
  • the device for collecting performance data of the network management system solves the network management system crash caused by a large number of thread concurrency by establishing a message queue including a performance data summary task and limiting the number of threads for performing parallel work of the performance data summary task. .
  • establishing the queue module 10 may include:
  • the classification unit 11 is configured to classify the message queue according to the category of the performance data summary task; the adding task unit 12 is configured to add the newly added performance data summary task to the corresponding message queue according to the category.
  • the classification unit 11 can classify the message queue according to certain characteristics of the performance data.
  • the measurement type of the performance data is taken as an example to classify the message queue.
  • the adding task unit 12 inserts the newly added performance data summary task into the tail of the corresponding message queue according to the measurement type, waiting for the thread to execute.
  • adding task unit 12 may include:
  • the determining sub-unit 121 is configured to determine a category of the newly added performance data summary task.
  • the first adding sub-unit 122 is configured to add a new performance data summary when the message queue of the category to which the new performance data summary task belongs is not full. Task to the message queue;
  • the second adding sub-unit 123 is configured to stop the enqueuing when the message queue of the category to which the new performance data summary task belongs is full, wait for the message queue to be taken away, create a new message queue, and add a new performance data summary. Task to the newly created message queue;
  • the third adding sub-unit 124 is configured to create a message queue when the message queue of the category to which the new performance data summary task belongs does not exist, and add a new performance data summary task to the newly created message queue.
  • the determining sub-unit 121 determines the category of the newly added performance data summary task. For example, when the performance data summary task of the measurement type CS Basic Measurement arrives, if the corresponding message queue is not full, the first adding sub-unit 122 summarizes the performance data task. Insert into the end of the corresponding message queue, waiting for the thread to execute.
  • the second adding sub-unit 123 stops the newly added performance data summary task from enqueuing, waits for the thread to take out the corresponding message queue, and then creates a new measurement type of CS Basic Measurement.
  • the message queue is added to the newly created message queue for the performance data summary task of CS Basic Measurement, which is waiting for and subsequently arriving.
  • the third adding sub-unit 124 creates a message queue of the measurement type CS Basic Measurement, and adds the performance data summary task of the measurement type CS Basic Measurement to the newly created one. In the message queue.
  • the method further includes:
  • the query module 30 is configured to periodically query whether the message queue is full.
  • the adjustment thread module 40 is configured to adjust the number of threads and/or the waiting time according to whether there is a full message queue, and the waiting time is a time interval at which the thread acquires the message queue.
  • the query module 30 can periodically query whether all message queues are full.
  • the query module 30 can implement the query function by setting a counter: when the message queue is full, the counter is incremented by 1. When the thread takes out a full message queue, the counter is decremented by 1, and the query module 30 can periodically obtain the counter count. If the counter count is non-zero, it means that at least one message queue is full. When the counter is 0, it means that all message queues are not full.
  • the query module 30 can also query by setting a full flag for each message queue, or by other means to achieve a full state of the query message queue.
  • the threading module 40 can reduce the waiting time of the thread and/or increase the number of threads working in parallel to speed up the execution speed of the performance data summary task, which is the time interval at which the thread acquires the message queue.
  • the adjustment thread module 40 can increase the waiting time of the thread and/or reduce the number of threads working in parallel to reduce the execution speed of the performance data summary task.
  • the adjustment thread module 40 includes:
  • the first adjusting unit 41 is configured to reduce a waiting time of the thread when there is a full message queue, and when the waiting time of the thread reaches a preset minimum value, create a new thread until the number of threads reaches a preset value;
  • the second adjusting unit 42 is configured to increase a waiting time of the thread when there is no full message queue, and end the thread when the waiting time of the thread reaches a preset maximum value.
  • the first adjusting unit 41 can adjust the number of threads and/or waiting time as follows: Query module 30 When querying If there is a full message queue, the first adjusting unit 41 reduces the waiting time of the thread in a certain order. If there are multiple threads, the waiting time of the first thread is reduced by 50 ms. When the query module 30 queries next time, it still exists.
  • the thread creation module 20 is triggered to create a new thread. The number of threads after the new thread cannot exceed the preset value, otherwise the new thread is stopped.
  • the second adjusting unit 42 can adjust the number of threads and/or waiting time in the following manner:
  • the query module 30 queries a certain time. When all message queues are not full, the waiting time of the thread is increased in a certain order. If there are multiple threads, the waiting time of the first thread is increased by 50 ms. When the query module 30 queries next time, all message queues are still not satisfied, and continue to increase. Waiting time of the first thread Between 50ms, until the waiting time of the first thread has reached the maximum value, the thread ends, and the second thread continues to be processed as above until there is only one thread, and the waiting time of the thread is the preset maximum value. If there is no message queue to be processed, the second adjustment unit 42 may end all threads.
  • the adjustment thread module 40 can adjust the number of threads and the waiting time separately or simultaneously, and other ways that can be implemented by those skilled in the art to adjust the number of threads according to the performance data aggregation task can be applied to the present invention.
  • the threading module 40 is adjusted.
  • the device for collecting the performance data of the network management system prevents the network management system from crashing due to a large number of thread concurrency by establishing a message queue including the performance data summary task and limiting the number of threads performing the performance data summary task. Further, the thread can be adjusted according to the number of performance data summary tasks to improve the performance of the network management system.

Landscapes

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

Abstract

本发明公开了一种网管系统性能数据汇总的方法和装置,其中方法包括:建立多个消息队列,所述消息队列包括性能数据汇总任务;建立数量不超过一预设值的线程,所述线程获取消息队列,并执行所述消息队列中性能数据汇总任务。本发明提出的网管系统性能数据汇总的方法和装置,通过建立包括性能数据汇总任务的消息队列,以及限制执行性能数据汇总任务的线程数量,防止了因大量线程并发造成的网管系统崩溃。

Description

网管系统性能数据汇总的方法和装置 技术领域
本发明涉及通信领域, 特别是指一种网管系统性能数据汇总的方法和 装置。 背景技术
性能管理是电信网络管理中的重要功能之一, 性能管理的目的是对网 络、 网络单元或设备进行性能监视, 釆集相关的性能统计数据, 评价网络 和网络单元的有效性, 报告电信设备的状态, 以及支持网络规划和网络分 析。 通常网管系统在釆集到原始的性能数据后, 需要将性能数据汇总为不 同时间粒度的数据存放, 如小时、 天、 周、 月等粒度, 以便查询时有直接 的数据提取, 提高网管系统效率。
性能数据汇总方法通常使用并行汇总, 传统的并行汇总方法是当网管 系统釆集到一批性能数据后, 便建立一批线程对该性能数据进行汇总。 上 述并行汇总的方法存在下述缺陷: 由于电信网络所具有的规模庞大、 高并 发性、 高实时性等特点, 当大量性能数据到达时, 网管系统需要建立大量 线程进行汇总, 极易产生并行风暴而导致网管系统崩溃。 发明内容
有鉴于此, 本发明的主要目的在于提供一种网管系统性能数据汇总的 方法和装置, 通过建立包括性能数据汇总任务的消息队列, 以及限制执行 性能数据汇总任务的线程数量, 防止因大量线程并发造成的网管系统崩溃。
为解决上述技术问题, 本发明的技术方案是这样实现的:
一种网管系统性能数据汇总的方法, 包括: 建立多个消息队列, 所述 消息队列包括性能数据汇总任务; 建立数量不超过一预设值的线程, 所述 线程获取消息队列, 并执行所述消息队列中性能数据汇总任务。
所述建立多个消息队列, 包括: 按性能数据汇总任务的类别对消息队 列进行分类; 将新增的性能数据汇总任务根据类别添加到对应的消息队列 中。
所述将新增的性能数据汇总任务根据类别添加到对应的消息队列中, 包括: 判断新增的性能数据汇总任务类别;
当新增的性能数据汇总任务所属类别的消息队列不满时, 添加新增的 性能数据汇总任务到该消息队列; 或,
当新增的性能数据汇总任务所属类别的消息队列已满时, 停止入队, 等待该消息队列取走后新建一消息队列, 并添加新增的性能数据汇总任务 到新建的消息队列; 或,
当新增的性能数据汇总任务所属类别的消息队列不存在时, 新建一消 息队列, 并添加新增的性能数据汇总任务到新建的消息队列。
在所述执行所述建立数量不超过一预设值的线程之后, 进一步包括: 定时查询所述消息队列是否已满; 根据是否存在已满的消息队列, 调整所 述线程的数量和 /或等待时间, 所述等待时间为线程获取消息队列的时间间 隔。
所述根据是否存在已满的消息队列调整线程的数量和 /或等待时间, 包 括:
当存在已满的消息队列时, 减少所述线程的等待时间, 当所述线程的 等待时间达到预设的最小值时, 新建一线程直到线程数量达到所述预设值; 当不存在已满的消息队列时, 增加所述线程的等待时间, 当所述线程 的等待时间达到预设的最大值时, 结束所述线程。
一种网管系统性能数据汇总的装置, 包括: 建立队列模块, 用于建立多个消息队列, 所述消息队列包括性能数据 汇总任务;
建立线程模块, 用于建立数量不超过一预设值的线程, 所述线程获取 消息队列, 用于执行所述消息队列中性能数据汇总任务。
所述建立队列模块包括:
分类单元, 用于按性能数据汇总任务的类别对消息队列进行分类; 添加任务单元, 用于将新增的性能数据汇总任务根据类别添加到对应 的消息队列中。
所述添加任务单元包括:
判断子单元, 用于判断新增的性能数据汇总任务的类别;
第一添加子单元, 用于当新增的性能数据汇总任务所属类别的消息队 列不满时, 添加新增的性能数据汇总任务到该消息队列;
第二添加子单元, 用于当新增的性能数据汇总任务所属类别的消息队 列已满时, 停止入队, 等待该消息队列取走后新建一消息队列, 并添加新 增的性能数据汇总任务到新建的消息队列;
第三添加子单元, 用于当新增的性能数据汇总任务所属类别的消息队 列不存在时, 新建一消息队列, 并添加新增的性能数据汇总任务到新建的 消息队列。
所述装置进一步包括:
查询模块, 用于定时查询所述消息队列是否已满;
调整线程模块, 用于根据是否存在已满的消息队列, 调整所述线程的 数量和 /或等待时间, 所述等待时间为线程获取消息队列的时间间隔。
所述调整线程模块包括:
第一调整单元, 用于当存在已满的消息队列时, 减少所述线程的等待 时间, 当所述线程的等待时间达到预设的最小值时, 触发所述建立线程模 块新建一线程直到线程数量达到所述预设值;
第二调整单元, 用于当不存在已满的消息队列时, 增加所述线程的等 待时间, 当所述线程的等待时间达到预设的最大值时, 结束所述线程。
本发明提供的一种网管系统性能数据汇总的方法和装置, 首先建立多 个包括性能数据汇总任务的消息队列, 再建立不超过一预设值的线程执行 性能数据汇总任务。 限制了并行工作的线程数量, 防止了因大量线程并发 造成的网管系统崩溃。 附图说明
图 1为本发明中网管系统性能数据汇总的方法一实施例的流程示意图; 图 2为本发明中网管系统性能数据汇总的方法一实施例中建立消息队 列的流程示意图;
图 3 为本发明中网管系统性能数据汇总的方法一实施例中添加新增的 性能数据汇总任务到消息队列的流程示意图;
图 4为本发明中网管系统性能数据汇总的方法又一实施例的流程示意 图;
图 5 为本发明中网管系统性能数据汇总的方法又一实施例中调整线程 的流程示意图;
图 6为本发明中网管系统性能数据汇总的装置一实施例的结构示意图; 图 7为本发明中网管系统性能数据汇总的装置一实施例中建立队列模 块的结构示意图;
图 8为本发明中网管系统性能数据汇总的装置一实施例中添加任务单 元的结构示意图;
图 9为本发明中网管系统性能数据汇总的装置又一实施例的结构示意 图;
图 10为本发明中网管系统性能数据汇总的装置又一实施例中调整线程 模块的结构示意图。 具体实施方式
本发明目的的实现、 功能特点及优点将结合实施例, 参照附图做进一 步说明。
本发明提供的一种网管系统性能数据汇总的方法和装置, 首先建立多 个包括性能数据汇总任务的消息队列, 再建立不超过一预设值的线程执行 性能数据汇总任务, 从而限制了并行工作的线程数量, 防止因大量线程并 发造成的网管系统崩溃。
参照图 1 , 提出本发明一种网管系统性能数据汇总的方法, 包括: 步骤 101 , 建立多个消息队列, 消息队列包括性能数据汇总任务。
网管系统首先建立多个消息队列, 消息队列包括性能数据汇总任务, 每一性能数据汇总任务包括网管系统所釆集的性能数据的信息, 例如性能 数据的时间和位置等。 消息队列之间为非线性的, 各个消息队列的长度可 以相同或不同, 可预先设置并根据需要进行调整, 以容纳一定数量的性能 数据汇总任务。
步骤 102, 建立数量不超过一预设值的线程, 线程获取消息队列, 并执 行该消息队列中的性能数据汇总任务。
网管系统接着建立执行性能数据汇总任务的线程, 线程的最大数量为 一预设值, 以防止大量线程并发时网管系统出现崩溃。 当有多个线程时, 各线程独立工作, 每一线程可定时或随机获取一个或多个待处理的消息队 歹 ij , 一般一线程一次处理一个待处理的消息队列, 然后依次执行该消息队 列中的性能数据汇总任务, 直到该消息队列中所有性能数据汇总任务执行 完毕, 最后线程可将该消息队列销毁, 再继续获取下一待处理的消息队列。 线程执行的性能数据汇总任务, 主要是将同一个粒度如一小时或一天的多 个性能数据汇总分析, 完成求和、 求平均、 求最大、 求最小等功能。 如上所述, 每一线程可根据能力一次处理一个或多个消息队列, 防止 各线程负担过重, 同时由于并行工作的线程数量为一预设值, 可防止大量 线程并发时网管系统出现崩溃。 上述并行工作的线程数量为网管系统预先 设置, 可根据网管系统能力调整。
本发明提供的一种网管系统性能数据汇总的方法, 通过建立包括性能 数据汇总任务的消息队列, 以及限制执行性能数据汇总任务的并行工作的 线程数量, 解决了因大量线程并发造成的网管系统崩溃。
参照图 2, 在一实施例中, 步骤 101可包括:
步骤 1011 , 按性能数据汇总任务的类别对消息队列进行分类。
网管系统可对消息队列按性能数据的一定特征进行分类, 本实施例以 性能数据的测量类型为例, 对消息队列进行分类。
步骤 1012 , 将新增的性能数据汇总任务根据类别添加到对应的消息队 列中。
当网管系统釆集到性能数据, 需要进行汇总时, 将新增的性能数据汇 总任务根据测量类型插入到对应的消息队列的队尾, 等待线程执行。
参照图 3 , 步骤 1012可包括:
步骤 1021 , 判断新增的性能数据汇总任务类别;
步骤 1022 , 当新增的性能数据汇总任务所属类别的消息队列不满时, 添加新增的性能数据汇总任务到该消息队列; 或,
步骤 1023 , 当新增的性能数据汇总任务所属类别的消息队列已满时, 停止入队, 等待该消息队列取走后新建一消息队列, 并添加新增的性能数 据汇总任务到新建的消息队列; 或,
步骤 1024,当新增的性能数据汇总任务所属类别的消息队列不存在时, 新建一消息队列, 并添加新增的性能数据汇总任务到新建的消息队列。
网管系统判断新增的性能数据汇总任务的类别,例如测量类型 CS Basic Measurement的性能数据汇总任务到达时, 如对应的消息队列不满, 将该性 能数据汇总任务插入到对应的消息队列的队尾, 等待线程执行。
当测量类型 CS Basic Measurement对应的消息队列已满时,停止新增的 性能数据汇总任务入队, 等待线程取出对应的消息队列后, 再新建一测量 类型为 CS Basic Measurement的消息队列,并将等待及后续到达的测量类型 为 CS Basic Measurement的性能数据汇总任务添加到新建的消息队列中。
当不存在测量类型 CS Basic Measurement对应的消息队列时,新建一测 量类型为 CS Basic Measurement的消息队列,将到达的测量类型为 CS Basic Measurement的性能数据汇总任务添加到新建的消息队列中。
参照图 4,提出本发明网管系统性能数据汇总的方法又一实施例, 在上 述实施例中, 在执行步骤 102之后, 还包括:
步骤 103 , 定时查询消息队列是否已满。
网管系统可定时查询所有消息队列是否已满。 网管系统可通过设置一 计数器来实现该查询功能: 当有消息队列已满时, 计数器加 1 , 当线程取出 一已满的消息队列时, 计数器减 1 , 网管系统可定时获取计数器的计数, 如 计数器的计数非 0, 代表存在至少一消息队列已满, 当计数器为 0, 则代表 所有消息队列不满。 网管系统也可通过对每一消息队列设置已满标识来查 询, 或釆用其它可实现查询消息队列已满状态的方式。
步骤 104, 根据是否存在已满的消息队列, 调整线程的数量和 /或等待 时间, 等待时间为线程获取消息队列的时间间隔。
当存在已满的消息队列时, 说明待执行的性能数据汇总任务较多, 网 管系统可相应减少线程的等待时间和 /或增加并行工作的线程数量, 以加快 性能数据汇总任务的执行速度, 线程的等待时间即线程获取消息队列的时 间间隔。 当所有消息队列均不满时, 说明待执行的性能数据汇总任务较少, 网管系统可相应增加线程的等待时间和 /或减少并行工作的线程数量, 以降 低性能数据汇总任务的执行速度。
参照图 5 , 步骤 104可包括:
步骤 1041 , 当存在已满的消息队列时, 减少线程的等待时间, 当线程 的等待时间达到预设的最小值时, 新建一线程直到线程数量达到预设值。
当存在已满的消息队列时, 说明消息队列中等待执行的性能数据汇总 任务较多, 可釆取如下方式调整线程的数量和 /或等待时间: 网管系统某次 查询时, 存在已满的消息队列, 则按一定顺序减少线程的等待时间, 如有 多个线程时, 减少第一个线程的等待时间 50ms, 当下次查询, 仍然存在已 满的消息队列时, 继续减少第一个线程的等待时间 50ms, 直到第一个线程 的等待时间已达最小值时, 继续对第二个线程作如上处理, 当所有线程的 等待时间都为最小值时, 新建一线程, 新建线程后的线程数量不能超过预 设值, 否则停止新建线程。
步骤 1042 , 当不存在已满的消息队列时, 增加线程的等待时间, 当线 程的等待时间达到预设的最大值时, 结束线程。
当不存在已满的消息队列时, 说明消息队列中等待执行的性能数据汇 总任务较少, 可釆取如下方式调整线程的数量和 /或等待时间: 网管系统某 次查询时, 所有消息队列不满, 则按一定顺序增加线程的等待时间, 如有 多个线程时, 增加第一个线程的等待时间 50ms, 当下次查询, 所有消息队 列仍然不满, 继续增加第一个线程的等待时间 50ms, 直到第一个线程的等 待时间已达最大值时, 结束该线程, 并继续对第二个线程作如上处理, 直 到只余一个线程, 且该线程的等待时间为预设的最大值。 如不存在待处理 的消息队列时, 可结束所有线程。
需要注意的是, 网管系统可以单独或同时调整线程数量和等待时间, 本领域的技术人员因此联想到的其它可实现根据性能数据汇总任务的数量 调整线程的方式均可适用于本发明。 本发明提供的一种网管系统性能数据汇总的方法, 通过建立包括性能 数据汇总任务的消息队列, 以及限制执行性能数据汇总任务的线程数量, 防止了因大量线程并发造成的网管系统崩溃。 进一步地, 可根据性能数据 汇总任务的数量调整线程, 提高网管系统性能。
参照图 6, 提出本发明网管系统性能数据汇总的装置一实施例, 包括: 建立队列模块 10, 用于建立多个消息队列, 消息队列包括性能数据汇 总任务;
建立线程模块 20, 用于建立数量不超过一预设值的线程, 线程获取消 息队列, 用于执行消息队列中的性能数据汇总任务。
建立队列模块 10首先建立多个消息队列, 消息队列包括性能数据汇总 任务, 每一性能数据汇总任务包括网管系统所釆集的性能数据的信息, 例 如性能数据的时间和位置等。 消息队列之间为非线性的, 各个消息队列的 长度可以相同或不同, 可预先设置并根据需要进行调整, 以容纳一定数量 的性能数据汇总任务。
建立线程模块 20接着建立执行性能数据汇总任务的线程, 线程的最大 数量为一预设值, 以防止大量线程并发时网管系统出现崩溃。 当有多个线 程时, 各线程独立工作, 每一线程可定时或随机获取一个或多个待处理的 消息队列, 一般一线程一次处理一个待处理的消息队列, 然后依次执行该 消息队列中的性能数据汇总任务, 直到该消息队列中所有性能数据汇总任 务执行完毕, 最后线程可将该消息队列销毁, 再继续获取下一待处理的消 息队列。 线程执行的性能数据汇总任务, 主要是将同一个粒度如一小时或 一天的多个性能数据汇总分析, 完成求和、 求平均、 求最大、 求最小等功 能。
如上所述, 每一线程可根据能力一次处理一个或多个消息队列, 防止 各线程负担过重, 同时由于并行工作的线程数量为一预设值, 可防止大量 线程并发时网管系统出现崩溃。 上述并行工作的线程数量为网管系统预先 设置, 可根据网管系统能力调整。
本发明提供的一种网管系统性能数据汇总的装置, 通过建立包括性能 数据汇总任务的消息队列, 以及限制执行性能数据汇总任务的并行工作的 线程数量, 解决了因大量线程并发造成的网管系统崩溃。
参照图 7, 在一实施例的网管系统性能数据汇总的装置中, 建立队列模 块 10可包括:
分类单元 11 , 用于按性能数据汇总任务的类别对消息队列进行分类; 添加任务单元 12 , 用于将新增的性能数据汇总任务根据类别添加到对 应的消息队列中。
分类单元 11可对消息队列按性能数据的一定特征进行分类, 本实施例 以性能数据的测量类型为例, 对消息队列进行分类。
当网管系统釆集到性能数据, 需要进行汇总时, 添加任务单元 12将新 增的性能数据汇总任务根据测量类型插入到对应的消息队列的队尾, 等待 线程执行。
参照图 8, 添加任务单元 12可包括:
判断子单元 121 , 用于判断新增的性能数据汇总任务的类别; 第一添加子单元 122 ,用于当新增的性能数据汇总任务所属类别的消息 队列不满时, 添加新增的性能数据汇总任务到该消息队列;
第二添加子单元 123 ,用于当新增的性能数据汇总任务所属类别的消息 队列已满时, 停止入队, 等待该消息队列取走后新建一消息队列, 并添加 新增的性能数据汇总任务到新建的消息队列;
第三添加子单元 124,用于当新增的性能数据汇总任务所属类别的消息 队列不存在时, 新建一消息队列, 并添加新增的性能数据汇总任务到新建 的消息队列。 例如, 判断子单元 121 判断新增的性能数据汇总任务的类别, 如测量 类型 CS Basic Measurement的性能数据汇总任务到达时,如对应的消息队列 不满, 第一添加子单元 122将该性能数据汇总任务插入到对应的消息队列 的队尾, 等待线程执行。
当测量类型 CS Basic Measurement对应的消息队列已满时,第二添加子 单元 123停止新增的性能数据汇总任务入队, 等待线程取出对应的消息队 列后,再新建一测量类型为 CS Basic Measurement的消息队列,并将等待及 后续到达的测量类型为 CS Basic Measurement的性能数据汇总任务添加到 新建的消息队列中。
当不存在测量类型 CS Basic Measurement对应的消息队列时,第三添加 子单元 124新建一测量类型为 CS Basic Measurement的消息队列,将到达的 测量类型为 CS Basic Measurement的性能数据汇总任务添加到新建的消息 队列中。
参照图 9,提出本发明网管系统性能数据汇总的装置又一实施例, 在上 述网管系统性能数据汇总的装置一实施例中, 还包括:
查询模块 30, 用于定时查询消息队列是否已满;
调整线程模块 40, 用于根据是否存在已满的消息队列, 调整线程的数 量和 /或等待时间, 该等待时间为线程获取消息队列的时间间隔。
查询模块 30可定时查询所有消息队列是否已满。 查询模块 30可通过 设置一计数器来实现该查询功能: 当有消息队列已满时, 计数器加 1 , 当线 程取出一已满的消息队列时, 计数器减 1 , 查询模块 30可定时获取计数器 的计数, 如计数器的计数非 0, 代表存在至少一消息队列已满, 当计数器为 0, 则代表所有消息队列不满。 查询模块 30也可通过对每一消息队列设置 已满标识来查询, 或釆用其它可实现查询消息队列已满状态的方式。
当存在已满的消息队列时, 说明待执行的性能数据汇总任务较多, 调 整线程模块 40可相应减少线程的等待时间和 /或增加并行工作的线程数量, 以加快性能数据汇总任务的执行速度, 线程的等待时间即线程获取消息队 列的时间间隔。 当所有消息队列均不满时, 说明待执行的性能数据汇总任 务较少,调整线程模块 40可相应增加线程的等待时间和 /或减少并行工作的 线程数量, 以降低性能数据汇总任务的执行速度。
参照图 10, 调整线程模块 40包括:
第一调整单元 41 , 用于当存在已满的消息队列时, 减少线程的等待时 间, 当线程的等待时间达到预设的最小值时, 新建一线程直到线程数量达 到预设值;
第二调整单元 42 , 用于当不存在已满的消息队列时, 增加线程的等待 时间, 当线程的等待时间达到预设的最大值时, 结束线程。
当存在已满的消息队列时, 说明消息队列中等待执行的性能数据汇总 任务较多, 第一调整单元 41 可釆取如下方式调整线程的数量和 /或等待时 间: 查询模块 30 某次查询时, 存在已满的消息队列, 则第一调整单元 41 按一定顺序减少线程的等待时间, 如有多个线程时, 减少第一个线程的等 待时间 50ms, 当查询模块 30下次查询, 仍然存在已满的消息队列时, 继 续减少第一个线程的等待时间 50ms, 直到第一个线程的等待时间已达最小 值时, 继续对第二个线程作如上处理, 当所有线程的等待时间都为最小值 时, 触发建立线程模块 20新建一线程, 新建线程后的线程数量不能超过预 设值, 否则停止新建线程。
当不存在已满的消息队列时, 说明消息队列中等待执行的性能数据汇 总任务较少,第二调整单元 42可釆取如下方式调整线程的数量和 /或等待时 间: 查询模块 30某次查询时, 所有消息队列不满, 则按一定顺序增加线程 的等待时间, 如有多个线程时, 增加第一个线程的等待时间 50ms, 当查询 模块 30下次查询, 所有消息队列仍然不满, 继续增加第一个线程的等待时 间 50ms, 直到第一个线程的等待时间已达最大值, 结束该线程, 并继续对 第二个线程作如上处理, 直到只余一个线程, 且该线程的等待时间为预设 的最大值。 如不存在待处理的消息队列时, 第二调整单元 42可结束所有线 程。
需要注意的是, 调整线程模块 40可以单独或同时调整线程数量和等待 时间, 本领域的技术人员因此联想到的其它可实现根据性能数据汇总任务 的数量调整线程的方式均可适用于本发明的调整线程模块 40。
本发明提供的一种网管系统性能数据汇总的装置, 通过建立包括性能 数据汇总任务的消息队列, 以及限制执行性能数据汇总任务的线程数量, 防止了因大量线程并发造成的网管系统崩溃。 进一步地, 可根据性能数据 汇总任务的数量调整线程, 提高网管系统性能。
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、 一种网管系统性能数据汇总的方法, 其特征在于, 包括:
建立多个消息队列, 所述消息队列包括性能数据汇总任务;
建立数量不超过一预设值的线程, 所述线程获取消息队列, 并执行所 述消息队列中性能数据汇总任务。
2、 如权利要求 1所述的网管系统性能数据汇总的方法, 其特征在于, 所述建立多个消息队列, 包括:
按性能数据汇总任务的类别对消息队列进行分类;
将新增的性能数据汇总任务根据类别添加到对应的消息队列中。
3、 如权利要求 2所述的网管系统性能数据汇总的方法, 其特征在于, 所述将新增的性能数据汇总任务根据类别添加到对应的消息队列中, 包括: 判断新增的性能数据汇总任务类别;
当新增的性能数据汇总任务所属类别的消息队列不满时, 添加新增的 性能数据汇总任务到该消息队列; 或,
当新增的性能数据汇总任务所属类别的消息队列已满时, 停止入队, 等待该消息队列取走后新建一消息队列, 并添加新增的性能数据汇总任务 到新建的消息队列; 或,
当新增的性能数据汇总任务所属类别的消息队列不存在时, 新建一消 息队列, 并添加新增的性能数据汇总任务到新建的消息队列。
4、 如权利要求 1至 3中任一项所述的网管系统性能数据汇总的方法, 其特征在于, 在所述执行所述建立数量不超过一预设值的线程之后, 进一 步包括:
定时查询所述消息队列是否已满;
根据是否存在已满的消息队列, 调整所述线程的数量和 /或等待时间, 所述等待时间为线程获取消息队列的时间间隔。
5、 如权利要求 4所述的网管系统性能数据汇总的方法, 其特征在于, 所述根据是否存在已满的消息队列调整线程的数量和 /或等待时间, 包括: 当存在已满的消息队列时, 减少所述线程的等待时间, 当所述线程的 等待时间达到预设的最小值时, 新建一线程直到线程数量达到所述预设值; 当不存在已满的消息队列时, 增加所述线程的等待时间, 当所述线程 的等待时间达到预设的最大值时, 结束所述线程。
6、 一种网管系统性能数据汇总的装置, 其特征在于, 包括:
建立队列模块, 用于建立多个消息队列, 所述消息队列包括性能数据 汇总任务;
建立线程模块, 用于建立数量不超过一预设值的线程, 所述线程获取 消息队列, 用于执行所述消息队列中性能数据汇总任务。
7、 如权利要求 6所述的网管系统性能数据汇总的装置, 其特征在于, 所述建立队列模块包括:
分类单元, 用于按性能数据汇总任务的类别对消息队列进行分类; 添加任务单元, 用于将新增的性能数据汇总任务根据类别添加到对应 的消息队列中。
8、 如权利要求 7所述的网管系统性能数据汇总的装置, 其特征在于, 所述添加任务单元包括:
判断子单元, 用于判断新增的性能数据汇总任务的类别;
第一添加子单元, 用于当新增的性能数据汇总任务所属类别的消息队 列不满时, 添加新增的性能数据汇总任务到该消息队列;
第二添加子单元, 用于当新增的性能数据汇总任务所属类别的消息队 列已满时, 停止入队, 等待该消息队列取走后新建一消息队列, 并添加新 增的性能数据汇总任务到新建的消息队列;
第三添加子单元, 用于当新增的性能数据汇总任务所属类别的消息队 列不存在时, 新建一消息队列, 并添加新增的性能数据汇总任务到新建的 消息队列。
9、 如权利要求 6至 8中任一项所述的网管系统性能数据汇总的装置, 其特征在于, 所述装置进一步包括:
查询模块, 用于定时查询所述消息队列是否已满;
调整线程模块, 用于根据是否存在已满的消息队列, 调整所述线程的 数量和 /或等待时间, 所述等待时间为线程获取消息队列的时间间隔。
10、 如权利要求 9所述的网管系统性能数据汇总的装置, 其特征在于, 所述调整线程模块包括:
第一调整单元, 用于当存在已满的消息队列时, 减少所述线程的等待 时间, 当所述线程的等待时间达到预设的最小值时, 触发所述建立线程模 块新建一线程直到线程数量达到所述预设值;
第二调整单元, 用于当不存在已满的消息队列时, 增加所述线程的等 待时间, 当所述线程的等待时间达到预设的最大值时, 结束所述线程。
PCT/CN2010/080112 2010-09-10 2010-12-22 网管系统性能数据汇总的方法和装置 WO2012031439A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN 201010279843 CN102402459B (zh) 2010-09-10 2010-09-10 网管系统性能数据汇总的方法和装置
CN201010279843.1 2010-09-10

Publications (1)

Publication Number Publication Date
WO2012031439A1 true WO2012031439A1 (zh) 2012-03-15

Family

ID=45810084

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/080112 WO2012031439A1 (zh) 2010-09-10 2010-12-22 网管系统性能数据汇总的方法和装置

Country Status (2)

Country Link
CN (1) CN102402459B (zh)
WO (1) WO2012031439A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108829521A (zh) * 2018-06-13 2018-11-16 平安科技(深圳)有限公司 任务处理方法、装置、计算机设备及存储介质

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6387747B2 (ja) * 2013-09-27 2018-09-12 日本電気株式会社 情報処理装置、障害回避方法およびコンピュータプログラム
CN104991814A (zh) * 2015-06-19 2015-10-21 北京奇虎科技有限公司 获取任务执行结果的方法及系统
CN105281962B (zh) * 2015-12-03 2018-08-28 成都广达新网科技股份有限公司 一种基于并行管道实现网管性能采集系统及其工作方法
CN105701257B (zh) * 2016-03-31 2019-05-21 北京奇虎科技有限公司 数据处理方法和装置
CN111221698A (zh) * 2018-11-26 2020-06-02 北京京东金融科技控股有限公司 任务数据采集方法与装置
CN110460534B (zh) * 2019-07-26 2024-05-14 腾讯云计算(北京)有限责任公司 一种请求消息上报方法、装置、设备及存储介质
CN111796930B (zh) * 2020-05-21 2023-12-22 福建天泉教育科技有限公司 优化消息处理的方法、存储介质
CN115314360B (zh) * 2022-07-27 2024-04-12 浪潮软件股份有限公司 消息队列服务性能监控系统及方法

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101262367A (zh) * 2008-03-07 2008-09-10 中兴通讯股份有限公司 性能数据的采集方法和装置
CN101420330A (zh) * 2008-12-11 2009-04-29 北京立通无限科技有限公司 邮件代理网关性能数据的采集方法和装置

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100347675C (zh) * 2004-06-29 2007-11-07 北京大学 应用服务器的性能优化方法
CN100518079C (zh) * 2004-09-30 2009-07-22 北京航空航天大学 分布式性能数据采集方法
CN1901568B (zh) * 2005-07-22 2010-09-01 中兴通讯股份有限公司 一种网管系统中实现历史性能采集的方法

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101262367A (zh) * 2008-03-07 2008-09-10 中兴通讯股份有限公司 性能数据的采集方法和装置
CN101420330A (zh) * 2008-12-11 2009-04-29 北京立通无限科技有限公司 邮件代理网关性能数据的采集方法和装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108829521A (zh) * 2018-06-13 2018-11-16 平安科技(深圳)有限公司 任务处理方法、装置、计算机设备及存储介质

Also Published As

Publication number Publication date
CN102402459A (zh) 2012-04-04
CN102402459B (zh) 2013-09-11

Similar Documents

Publication Publication Date Title
WO2012031439A1 (zh) 网管系统性能数据汇总的方法和装置
JP5233504B2 (ja) 経路制御装置およびパケット廃棄方法
CN102088363B (zh) 告警处理方法和系统
US20110196964A1 (en) Managing event traffic in a network system
US10295368B2 (en) Data collection method, apparatus, and system
CN106878132B (zh) 基于周期及优先级的总线调度方法
EP2426871B1 (en) Method and device for scheduling data communication input ports
DE60308672T2 (de) Flexible architektur zur planung von warteschlangen in einem paketvermittelten netz
CN102970244A (zh) 一种多cpu核间负载均衡的网络报文处理方法
WO2004028083A1 (fr) Procede d'ordonnancement pour les donnees de dispositif d'invitation a emettre
CN107220123A (zh) 一种解决Spark数据倾斜方法及系统
CN102831046A (zh) 通讯系统中一种软件性能采集的方法及装置
DE112006002912T5 (de) Verwaltung von On-Chip-Warteschleifen in geschalteten Netzwerken
CN108512727A (zh) 一种中央处理器利用率的确定方法及装置
CN108809752B (zh) 一种网络流量的自适应监控方法、装置、npb设备及介质
CN107479966B (zh) 一种基于多核心cpu的信令采集方法
CN101442439B (zh) 一种上报中断的方法和pci总线系统
WO2022100312A1 (zh) 提升对Flash通道控制器消息处理效率的方法及装置
CN107332785A (zh) 一种基于动态工作阈值的有效流量控制方法
CN106775984B (zh) 一种管理线程池的方法和装置
CN101499935B (zh) 一种WiMAX基站的告警处理方法
WO2012079352A1 (zh) 一种多个网元设备的管理方法和系统
CN106412341A (zh) 一种通知信息推送方法、装置及终端
CN105610633A (zh) 一种通信设备中实时性能自采样系统及方法
WO2011131114A1 (zh) 一种网络处理器异常处理方法及系统

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: 10856901

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 10856901

Country of ref document: EP

Kind code of ref document: A1