zabbix_agentd断断续续端无法访问问题记录

问题现象：

zabbix监控上出现zabbix_agentd无法访问，但是实际上zabbix_agentd是存活状态

每隔一段时间就会出现这样的情况

问题原因

zabbix_agentd端任务较多，活动线程被占满，无法及时响应server端的确认存活信息，导致server端认为agent端未启动

解决方法

在配置文件中修改zabbix_agentd端的活动线程数（默认是3）,增加如下配置：

StartAgents=30

~]# vim /usr/local/zabbix_agent/etc/zabbix_agentd.conf
PidFile=/usr/local/zabbix_agent/pids/zabbix_agentd.pid
LogFile=/usr/local/zabbix_agent/logs/zabbix_agentd.log
LogFileSize=1
DebugLevel=3
Timeout=30
StartAgents=20

结果

默认未修改前，线程数如下，只有三条

~]# ps -ef|grep zabbix
zabbix     1404      1  0 Aug26 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd
zabbix     1406   1404  0 Aug26 ?        00:08:43 /usr/local/zabbix_agent/sbin/zabbix_agentd: collector [idle 1 sec]
zabbix     1407   1404  0 Aug26 ?        00:25:04 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #1 [processing request]
zabbix     1408   1404  0 Aug26 ?        00:24:07 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #2 [waiting for connection]
zabbix     1409   1404  0 Aug26 ?        00:24:33 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #3 [processing request]
zabbix     1410   1404  0 Aug26 ?        00:01:56 /usr/local/zabbix_agent/sbin/zabbix_agentd: active checks #1 [idle 1 sec]

修改后，线程变成了

~]# ps -ef|grep zabbix
zabbix   117198      1  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd
zabbix   117200 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: collector [idle 1 sec]
zabbix   117201 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #1 [waiting for connection]
zabbix   117202 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #2 [waiting for connection]
zabbix   117203 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #3 [waiting for connection]
zabbix   117204 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #4 [waiting for connection]
zabbix   117205 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #5 [waiting for connection]
zabbix   117206 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #6 [waiting for connection]
zabbix   117207 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #7 [waiting for connection]
zabbix   117208 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #8 [waiting for connection]
zabbix   117209 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #9 [waiting for connection]
zabbix   117210 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #10 [waiting for connection]
zabbix   117211 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #11 [waiting for connection]
zabbix   117212 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #12 [waiting for connection]
zabbix   117213 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #13 [waiting for connection]
zabbix   117214 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #14 [waiting for connection]
zabbix   117215 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #15 [waiting for connection]
zabbix   117216 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #16 [waiting for connection]
zabbix   117217 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #17 [waiting for connection]
zabbix   117218 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #18 [waiting for connection]
zabbix   117219 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #19 [waiting for connection]
zabbix   117220 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: listener #20 [waiting for connection]
zabbix   117221 117198  0 17:38 ?        00:00:00 /usr/local/zabbix_agent/sbin/zabbix_agentd: active checks #1 [idle 1 sec]
root     117223 112890  0 17:39 pts/0    00:00:00 grep --color=auto zabbix

根据实际生产环境对活动线程数进行调整就好