• 通达OA系统优化-对mysql数据库减肥


    OA系统冗余数据过多,访问效率受到影响,现需要对历史数据进行一次清理,以提高OA访问速度


      大的数据主要体现在流程上,流程数据主要放在flow_run,flow_run_data,flow_run_prcs,flow_run_feedback表中,清理一年半以前的数据是比较合理的。run_id是主键而run_id随着时间的增加而变大,run_id小于50W的流程刚好是一年半以前的历史数据,选择run_id小于50W作为临界点进行删除。以下是清理数据的计划


    思路:

    ①创建和flow_run_data、flow_run、flow_run_prcs、flow_run_feedback、notify表结构一致的表

    ②把需要清理的历史数据插入刚创建的备份表中

    ③删除正式表中的数据

    操作:

    1.晚上10点以后停止mysql数据库服务,前端web服务(避免有人继续访问)

    2.备份td_oa数据库

    3.依次执行以下SQL语句:

    4.开启web和mysql服务,验证是否成功


    //对flow_run的操作

    create table flow_run_old like flow_run;

    insert into flow_run_old select * from flow_run where run_id < '500000';

    delete from flow_run where run_id < '500000';

    //对flow_run_data的操作

    create table flow_run_data_old like flow_run_data;

    insert flow_run_data_old select * from flow_run_data where run_id < '500000'

    delete from flow_run_data where run_id < '500000';

    //对flow_run_prcs操作

    create table flow_run_prcs_old like flow_run_prcs;

    insert into flow_run_prcs_old select * from flow_run_prcs where run_id < '500000'

    delete from flow_run_prcs where run_id < '500000';

    select count(*) from flow_run_prcs_old

    //对flow_run_feedback操作

    create table flow_run_feedback_old like flow_run_feedback;

    insert into flow_run_feedback_old select * from flow_run_feedback where run_id < '500000'

    delete from flow_run_feedback where run_id < '500000';

    //公告notify_id<3349是2013年以前的数据,从该时刻开始删除

    create table notify_old like notify;

    insert into notify_old select * from notify where notify_id < '3349';

    delete from notify where notify_id < '3349';

    //删除一年以前的系统日志sys_log、流程日志FLOW_RUN_LOG、SMS、SMS_BODY、UC_PMS、

    delete from sys_log where time<date_sub(date(now()),interval 1 year);


    //对流程日志的操作

    create table flow_run_log_old like flow_run_log;

    insert into flow_run_log_old select * from flow_run_log where run_id < '500000';

    delete from flow_run_log where run_id < '500000';

    =============================================================================================================

    //会议室申请

    create table meeting_old like meeting;

    insert into meeting_old select * from meeting where m_request_time < '2013-06-01 00:00:00';

    delete from meeting where m_request_time < '2013-06-01 00:00:00';


    //车辆申请

    create table vehicle_usage_old like vehicle_usage;

    insert into vehicle_usage_old select * from vehicle_usage where vu_request_date < '2013-06-01 00:00:00';

    delete from vehicle_usage where vu_request_date < '2013-06-01 00:00:00';

    =============================================================================================================

    //在线考试信息清理主要是exam_flow和exam_data两张表,这两张表是通过flow_id关联,清理2012-06-01以前的数据,flow_id<4212

    //考试发布表exam_flow

    create table exam_flow_old like exam_flow;

    insert into exam_flow_old select * from exam_flow where flow_id<'4212';

    delete from exam_flow where flow_id<'4212';

    //考试信息表exam_data

    create table exam_data_old like exam_data;

    insert into exam_data_old select * from exam_data where flow_id<'4212';

    delete from exam_data where flow_id<'4212';

  • 相关阅读:
    DFS(White-Gray-Black)
    Recursive Depth first search graph(adj matrix)
    TED_Topic1:Why we need to rethink capitalism
    Eng1—English daily notes
    R2—《R in Nutshell》 读书笔记(连载)
    R1(上)—R关联规则分析之Arules包详解
    R1(下)—数据挖掘—关联规则理论介绍与R实现
    R0—New packages for reading data into R — fast
    Tools0—Word2013发布博客到博客园的详细方法介绍
    day6 类
  • 原文地址:https://www.cnblogs.com/reblue520/p/6239896.html
Copyright © 2020-2023  润新知