Python之路,Day20 - 分布式监控系统开发

本节内容

为什么要做监控？

常用监控系统设计讨论

监控系统架构设计

监控表结构设计

为什么要做监控？

–熟悉IT监控系统的设计原理

–开发一个简版的类Zabbix监控系统

–掌握自动化开发项目的程序设计思路及架构解藕原则

常用监控系统设计讨论

Zabbix

Nagios

监控系统需求讨论

1.可监控常用系统服务、应用、网络设备等

2.一台主机上可监控多个不同服务、不同服务的监控间隔可不同

3.同一个服务在不同主机上的监控间隔、报警阈值可不同

4.可以批量的给一批主机添加、删除、修改要监控的服务

5.告警级别：

不同的服务因为业务重要程度不同，如果出了问题可以设置不同的报警级别
可以指定特定的服务或告警级别的事件通知给特定的用户
告警的升级设定

6.历史数据的存储和优化

实现用最少的空间占用量存储最多的有效数据
如何做到1s中之内取出一台主机上所有服务的5年的监控数据？

7. 数据可视化，如何做出简洁美观的用户界面？

8.如何实现单机支持5000+机器监控需求？

9.采取何种通信方式？主动、被动？

10.如何实现监控服务器的水平扩展？

采用什么架构？

•Mysql

•主动通信？ Snmp,wget…

•被动通信？Agent ---how to communicate with the monitor server

•Socket server –> Sockect client

•能否用现成的c/s架构？ Rabbit mq, redis 订阅发布， http ?

采用HTTP好处

1.接口设计简单

2.容易水平扩展做分布式

3.Socket稳定成熟，省去较多的通信维护精力

Http特性：

1.短连接

2.无状态

3.安全认证

4.被动通信

监控系统架构设计

表结构设计

  1 #!_*_coding:utf8_*_

  2 from django.db import models

  3

  4 # Create your models here.

  5

  6

  7

  8

  9 class Host(models.Model):

 10     name =  models.CharField(max_length=64,unique=True)

 11     ip_addr =  models.GenericIPAddressField(unique=True)

 12     host_groups = models.ManyToManyField('HostGroup',blank=True) # A B C

 13     templates = models.ManyToManyField("Template",blank=True) # A D E

 14     monitored_by_choices = (

 15         ('agent','Agent'),

 16         ('snmp','SNMP'),

 17         ('wget','WGET'),

 18     )

 19     monitored_by = models.CharField(u'监控方式',max_length=64,choices=monitored_by_choices)

 20     status_choices= (

 21         (1,'Online'),

 22         (2,'Down'),

 23         (3,'Unreachable'),

 24         (4,'Offline'),

 25     )

 26     status = models.IntegerField(u'状态',choices=status_choices,default=1)

 27     memo = models.TextField(u"备注",blank=True,null=True)

 28

 29     def __unicode__(self):

 30         return self.name

 31

 32 class HostGroup(models.Model):

 33     name =  models.CharField(max_length=64,unique=True)

 34     templates = models.ManyToManyField("Template",blank=True)

 35     memo = models.TextField(u"备注",blank=True,null=True)

 36     def __unicode__(self):

 37         return self.name

 38

 39 class ServiceIndex(models.Model):

 40     name = models.CharField(max_length=64)

 41     key =models.CharField(max_length=64)

 42     data_type_choices = (

 43         ('int',"int"),

 44         ('float',"float"),

 45         ('str',"string")

 46     )

 47     data_type = models.CharField(u'指标数据类型',max_length=32,choices=data_type_choices,default='int')

 48     memo = models.CharField(u"备注",max_length=128,blank=True,null=True)

 49     def __unicode__(self):

 50         return "%s.%s" %(self.name,self.key)

 51

 52 class Service(models.Model):

 53     name = models.CharField(u'服务名称',max_length=64,unique=True)

 54     interval = models.IntegerField(u'监控间隔',default=60)

 55     plugin_name = models.CharField(u'插件名',max_length=64,default='n/a')

 56     items = models.ManyToManyField('ServiceIndex',verbose_name=u"指标列表",blank=True)

 57     memo = models.CharField(u"备注",max_length=128,blank=True,null=True)

 58

 59     def __unicode__(self):

 60         return self.name

 61     #def get_service_items(obj):

 62     #    return ",".join([i.name for i in obj.items.all()])

 63

 64 class Template(models.Model):

 65     name = models.CharField(u'模版名称',max_length=64,unique=True)

 66     services = models.ManyToManyField('Service',verbose_name=u"服务列表")

 67     triggers = models.ManyToManyField('Trigger',verbose_name=u"触发器列表",blank=True)

 68     def __unicode__(self):

 69         return self.name

 70 '''

 71 class TriggerExpression(models.Model):

 72     name = models.CharField(u"触发器表达式名称",max_length=64,blank=True,null=True)

 73     service = models.ForeignKey(Service,verbose_name=u"关联服务")

 74     service_index = models.ForeignKey(ServiceIndex,verbose_name=u"关联服务指标")

 75     logic_type_choices = (('or','OR'),('and','AND'))

 76     logic_type = models.CharField(u"逻辑关系",choices=logic_type_choices,max_length=32,blank=True,null=True)

 77     left_sibling = models.ForeignKey('self',verbose_name=u"左边条件",blank=True,null=True,related_name='left_sibling_condition' )

 78     operator_type_choices = (('eq','='),('lt','<'),('gt','>'))

 79     operator_type = models.CharField(u"运算符",choices=operator_type_choices,max_length=32)

 80     data_calc_type_choices = (

 81         ('avg','Average'),

 82         ('max','Max'),

 83         ('hit','Hit'),

 84         ('last','Last'),

 85     )

 86     data_calc_func= models.CharField(u"数据处理方式",choices=data_calc_type_choices,max_length=64)

 87     data_calc_args = models.CharField(u"函数传入参数",help_text=u"若是多个参数,则用,号分开,第一个值是时间",max_length=64)

 88     threshold = models.IntegerField(u"阈值")

 89

 90     def __unicode__(self):

 91         return "%s %s(%s(%s))" %(self.service_index,self.operator_type,self.data_calc_func,self.data_calc_args)

 92 '''

 93

 94

 95 class TriggerExpression(models.Model):

 96     #name = models.CharField(u"触发器表达式名称",max_length=64,blank=True,null=True)

 97     trigger = models.ForeignKey('Trigger',verbose_name=u"所属触发器")

 98     service = models.ForeignKey(Service,verbose_name=u"关联服务")

 99     service_index = models.ForeignKey(ServiceIndex,verbose_name=u"关联服务指标")

100     specified_index_key = models.CharField(verbose_name=u"只监控专门指定的指标key",max_length=64,blank=True,null=True)

101     operator_type_choices = (('eq','='),('lt','<'),('gt','>'))

102     operator_type = models.CharField(u"运算符",choices=operator_type_choices,max_length=32)

103     data_calc_type_choices = (

104         ('avg','Average'),

105         ('max','Max'),

106         ('hit','Hit'),

107         ('last','Last'),

108     )

109     data_calc_func= models.CharField(u"数据处理方式",choices=data_calc_type_choices,max_length=64)

110     data_calc_args = models.CharField(u"函数传入参数",help_text=u"若是多个参数,则用,号分开,第一个值是时间",max_length=64)

111     threshold = models.IntegerField(u"阈值")

112

113

114     logic_type_choices = (('or','OR'),('and','AND'))

115     logic_type = models.CharField(u"与一个条件的逻辑关系",choices=logic_type_choices,max_length=32,blank=True,null=True)

116     #next_condition = models.ForeignKey('self',verbose_name=u"右边条件",blank=True,null=True,related_name='right_sibling_condition' )

117     def __unicode__(self):

118         return "%s %s(%s(%s))" %(self.service_index,self.operator_type,self.data_calc_func,self.data_calc_args)

119     class Meta:

120         pass #unique_together = ('trigger_id','service')

121

122 class Trigger(models.Model):

123     name = models.CharField(u'触发器名称',max_length=64)

124     #expressions= models.TextField(u"表达式")

125     severity_choices = (

126         (1,'Information'),

127         (2,'Warning'),

128         (3,'Average'),

129         (4,'High'),

130         (5,'Diaster'),

131     )

132     #expressions = models.ManyToManyField(TriggerExpression,verbose_name=u"条件表达式")

133     severity = models.IntegerField(u'告警级别',choices=severity_choices)

134     enabled = models.BooleanField(default=True)

135     memo = models.TextField(u"备注",blank=True,null=True)

136

137     def __unicode__(self):

138         return "<serice:%s, severity:%s>" %(self.name,self.get_severity_display())

139

140

141

142 class Action(models.Model):

143     name =  models.CharField(max_length=64,unique=True)

144     host_groups = models.ManyToManyField('HostGroup',blank=True)

145     hosts = models.ManyToManyField('Host',blank=True)

146

147     conditions = models.TextField(u'告警条件')

148     interval = models.IntegerField(u'告警间隔(s)',default=300)

149     operations = models.ManyToManyField('ActionOperation')

150

151     recover_notice = models.BooleanField(u'故障恢复后发送通知消息',default=True)

152     recover_subject = models.CharField(max_length=128,blank=True,null=True)

153     recover_message = models.TextField(blank=True,null=True)

154

155     enabled = models.BooleanField(default=True)

156

157     def __unicode__(self):

158         return self.name

159

160 class ActionOperation(models.Model):

161     name =  models.CharField(max_length=64)

162     step = models.SmallIntegerField(u"第n次告警",default=1)

163     action_type_choices = (

164         ('email','Email'),

165         ('sms','SMS'),

166         ('script','RunScript'),

167     )

168     action_type = models.CharField(u"动作类型",choices=action_type_choices,default='email',max_length=64)

169     #notifiers= models.ManyToManyField(host_models.UserProfile,verbose_name=u"通知对象",blank=True)

170     def __unicode__(self):

171         return self.name

172

173

174 class Maintenance(models.Model):

175     name =  models.CharField(max_length=64,unique=True)

176     hosts = models.ManyToManyField('Host',blank=True)

177     host_groups = models.ManyToManyField('HostGroup',blank=True)

178     content = models.TextField(u"维护内容")

179     start_time = models.DateTimeField()

180     end_time = models.DateTimeField()

181

182     def __unicode__(self):

183         return self.name

184

185 ''''

186 CPU

187     idle 80

188     usage  90

189     system  30

190     user

191     iowait  50

192

193 memory :

194     usage

195     free

196     swap

197     cache

198     buffer

199

200 load:

201     load1

202     load 5

203     load 15

204 '''

分类: Python自动化开发之路

好文要顶关注我收藏该文

金角大王
关注 - 0
粉丝 - 639

+加关注

«上一篇：Python Select 解析

posted @ 2016-05-01 17:18 金角大王阅读(901) 评论(0) 编辑收藏

刷新评论刷新页面返回顶部