首页> 外国专利> FAULT PREDICTION MODEL BASED AGGRESSIVE CHECKPOINTING METHOD

FAULT PREDICTION MODEL BASED AGGRESSIVE CHECKPOINTING METHOD

机译:基于故障预测模型的激进检查点方法

摘要

The present invention relates to a check pointing method based on defect prediction. step; (b) The defect occurrence prediction time point provided by the predetermined defect occurrence prediction model is divided by the checkpoint interval and the total Determining the number of check points; (c) the entire check using the system error rate of the defect occurrence prediction model Calculating the number of checkpoints to be performed in the section before the occurrence of the defect from the number of points; (d) all check points above Checkpoints to be performed in the defective section by excluding the number of checkpoints to be performed in the section before the defect occurrence from the number Calculating the number of twitches; (e) Schedule according to the number of checkpoints calculated in step (c) above in the section prior to the occurrence of the defect. Storing a system snapshot by performing a checkpoint at an interval; (f) In the defect generation section, Storing a system snapshot by performing checkpoints at regular intervals according to the calculated number of checkpoints; And (g) In the event of a system failure, the step of rolling back the system to a time point closest to the time point of the failure using the system snapshot; It is characterized by including.
机译:基于缺陷预测的检查指向方法技术领域本发明涉及基于缺陷预测的检查指向方法。步; (b)由预定缺陷发生预测模型提供的缺陷发生预测时间点除以检查点间隔,并确定检查点的总数; (c)整个检查使用缺陷发生预测模型的系统错误率计算要在从点数的缺陷发生之前进行的检查点的数量; (d)通过排除在从计算抽搐数量的缺陷之前在缺陷之前进行的检查点的数量排除在缺陷部分中进行检查点以上的检查点。 (e)根据在缺陷发生之前的部分中的步骤(c)中计算的检查点数量的时间表。通过以间隔执行检查点来存储系统快照; (f)在缺陷生成部分中,通过根据计算的检查点的数量以规则的间隔执行检查点来存储系统快照; (g)如果系统发生故障,将系统滚动到最接近故障时间点的时间点使用系统快照;它的特征在于包括。

著录项

  • 公开/公告号KR20210026387A

    专利类型

  • 公开/公告日2021-03-10

    原文格式PDF

  • 申请/专利权人 주식회사 세찬아이앤씨;

    申请/专利号KR1020190107084

  • 发明设计人 전영걸;

    申请日2019-08-30

  • 分类号G06F11/36;G06F11/14;

  • 国家 KR

  • 入库时间 2022-08-24 17:36:06

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号