增删改查,最注重要是删和改。

1.主机表，应用表，主机应用关系表

主机表

DROP TABLE IF EXISTS `a`;
CREATE TABLE `a` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `ip` varchar(15) COLLATE utf8mb4_unicode_ci NOT NULL COMMENT 'ip地址',
  `env` varchar(15) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '环境',
  `system` varchar(40) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '系统类型',
  `application` varchar(255) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '应用',
  `scu` varchar(255) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '具体用途',
  PRIMARY KEY (`id`),
  UNIQUE KEY `ip` (`ip`) USING BTREE
) ENGINE=InnoDB AUTO_INCREMENT=72 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='最先存入的所有主机表';

主机应用关系表

DROP TABLE IF EXISTS `b`;
CREATE TABLE `b` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `ipid` int(11) DEFAULT NULL,
  `applyid` int(11) DEFAULT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=916 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;

应用表

DROP TABLE IF EXISTS `pub_datadictionary`;
CREATE TABLE `pub_datadictionary` (
  `dd_no` varchar(38) COLLATE utf8mb4_unicode_ci NOT NULL,
  `dd_code` varchar(100) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '字典项代码',
  `dd_name` varchar(100) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '字典项名称',
  `dd_desc` varchar(200) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '描述',
  `is_lock` char(1) COLLATE utf8mb4_unicode_ci DEFAULT 'N' COMMENT '是否锁定:Y-锁定,N未锁定',
  PRIMARY KEY (`dd_no`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;

DROP TABLE IF EXISTS `pub_dditem`;
CREATE TABLE `pub_dditem` (
  `item_no` varchar(38) COLLATE utf8mb4_unicode_ci NOT NULL,
  `item_code` varchar(100) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '常量值',
  `item_name` varchar(255) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT '常量名称',
  `item_order` int(11) DEFAULT '999' COMMENT '排序值',
  `dd_no` varchar(38) COLLATE utf8mb4_unicode_ci NOT NULL COMMENT '所属数据字典项',
  `is_lock` char(1) COLLATE utf8mb4_unicode_ci DEFAULT 'N' COMMENT '是否锁定:Y-锁定,N未锁定',
  `is_sync` varchar(1) COLLATE utf8mb4_unicode_ci DEFAULT 'N' COMMENT '是否同步',
  `parentno` varchar(255) COLLATE utf8mb4_unicode_ci DEFAULT NULL,
  `itemdesc` varchar(255) COLLATE utf8mb4_unicode_ci DEFAULT NULL,
  PRIMARY KEY (`item_no`),
  KEY `IDX_160_ITEMCODE` (`item_code`) USING BTREE
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;

查询的sql

SELECT
    d.item_name
FROM
    (
        SELECT
            b.applyid
        FROM
            b
        JOIN a ON b.ipid = a.id
        WHERE
            a.ip = '172.x.x.x'
    ) AS c
JOIN pub_dditem d ON c.applyid = d.item_no;

a
1 172.1.1.1 test Cloud mysql 数据库
2 172.1.1.2 test Cloud tomcat a+b+c,jenkins

b
空

pub_dditem
148 mysql mysql 46 100 N N 0 
149 tomcat a 47 101 N N 0
150 tomcat a 48 101 N N 0
151 tomcat a 49 101 N N 0

update_hosts_apply.xls
172.1.1.1 mysql
172.1.1.2 tomcat(crm+workflow+service)

python小脚本初始化（仅仅是插入）

# !/usr/bin/env python
# coding:utf-8
import json
import urllib2
from urllib2 import URLError
import sys
import xlrd
import MySQLdb as mdb

def conn(ip,user,password,database):
    con = mdb.connect(ip,user,password,database)
    cur = con.cursor()
    return cur

def select(cur,sql):
    cur.execute(sql)
    return cur.fetchall()

def insert(cur,xls):
    workbook = xlrd.open_workbook(xls)
    for row in xrange(workbook.sheets()[0].nrows):
        ip = workbook.sheets()[0].cell(row, 0).value
        applys = workbook.sheets()[0].cell(row, 1).value
        for apply in applys.split(','):
            if 'tomcat' in apply:
                tomcat_apply = apply.split('(')[1].split(')')[0].split('+')
                for ztomcat_apply in tomcat_apply:
                    sql = "select item_no from pub_dditem where item_code='tomcat' and item_name='%s' " % (
                    ztomcat_apply)
                    sql1 = "select id from a where ip='%s'" % (ip)
                    for r in select(cur, sql):
                        item_no = r[0]
                    for r in select(cur, sql1):
                        id = r[0]
                    sql2 = "insert into  b (ipid,applyid) values(%d,%d)" % (int(id), int(item_no))
                    sql3 = "select * from b where ipid = '%s' and applyid = '%s'" % (id, item_no)
                    flag = cur.execute(sql3)
                    if flag == 0L:
                        cur.execute(sql2)
                        con.commit()

            else:
                sql = "select item_no from pub_dditem where item_name='%s' " % (apply)
                sql1 = "select id from a where ip='%s'" % (ip)
                for r in select(cur, sql):
                    item_no = r[0]
                for r in select(cur, sql1):
                    id = r[0]
                sql2 = "insert into b (ipid,applyid) values(%d,%d)" % (int(id), int(item_no))
                sql3 = "select * from b where ipid = '%s' and applyid = '%s'" % (id, item_no)
                flag = cur.execute(sql3)
                if flag == 0L:
                    cur.execute(sql2)
                    con.commit()

if __name__ == "__main__":
    con = mdb.connect('db', 'root', 'password', 'cmdb')
    cur = con.cursor()
    insert(cur,'update_hosts_apply.xls')
    cur.close()

还有的问题:
1.初始化数据一定要准确，不然执行到b表会重复
2.录入的xls也要准确,比如出现写错应用的，那么b表也会重复

本文转自 liqius 51CTO博客，原文链接：http://blog.51cto.com/szgb17/2074648，如需转载请自行联系原作者