nagios报警延迟的解决--flapping state

转载请注明出处：blog.csdn.net/abnerfrmh

这个问题是在测试中发现的。因为要在服务器上布置nagios用来监控oracle，可是发现手动shutdown数据库后能够很快报警，但是再startup后就不是很及时，有时会延迟很久。经过研究发现了这个情况，现贴出来与大家分享。


注意看那个类似云朵的图标，那个就是在提示状态抖动，也就是flapping state检测开启了。而一旦开启，就会temporarily suppres notifications。

报警延迟是由于host or service状态不稳定，nagios称之为flapping state
在nagios的主配置文件nagios.cfg中，有如下内容：
# FLAP DETECTION OPTION
# This option determines whether or not Nagios will try
# and detect hosts and services that are "flapping".  
# Flapping occurs when a host or service changes between
# states too frequently.  When Nagios detects that a 
# host or service is flapping, it will temporarily suppress
# notifications for that host/service until it stops
# flapping.  Flap detection is very experimental, so read
# the HTML documentation before enabling this feature!
# Values: 1 = enable flap detection
#         0 = disable flap detection (default)


enable_flap_detection=0#虽然文档里说默认是0，但实际为1.改为0之后就解决了flapping state，并且解决了临时suppress notifications的问题






# FLAP DETECTION THRESHOLDS FOR HOSTS AND SERVICES
# Read the HTML documentation on flap detection for
# an explanation of what this option does.  This option
# has no effect if flap detection is disabled.


low_service_flap_threshold=5.0
high_service_flap_threshold=20.0
low_host_flap_threshold=5.0
high_host_flap_threshold=20.0