MYSQl Gone AWAY ??

[kloczek]

I had the same issue: Increasing following limits helped:
-/etc/my.cnf-
max_connections=1000 <-- higher then default
wait_timeouts=45000 <-- was 28800


- /etc/zabbix/zabbix_server.conf-

Timeout=30
StartPollers=500

restart mysqld
restart zabbix-server

You really need stop using passive monitoring and switch to active one.
Passive monitoring does not scales up.

[mkash28]

thanks kloczek,

in past when i attempted to configure agent for active checks, I ran in to issues. What should I look for in logs to determine that Active checks are happening successfully?

this is how agent log looks like (log-debug-level=4)

44232:20160408:190848.467 End of send_buffer():SUCCEED
44232:20160408:190848.467 active checks #1 [getting list of active checks]
44232:20160408:190848.467 In refresh_active_checks() host:’*.*.*.*’ port:10051
44229:20160408:190849.000 listener #1 [processing request]
44229:20160408:190849.000 Requested [vfs.fs.size[/archive/ease2g/source,total]]
44229:20160408:190849.000 listener #1 [waiting for connection]
44228:20160408:190849.112 collector [processing data]
44228:20160408:190849.112 In update_cpustats()
44228:20160408:190849.113 End of update_cpustats()
44228:20160408:190849.113 collector [idle 1 sec]
44229:20160408:190850.005 listener #1 [processing request]
44229:20160408:190850.005 Requested [vfs.fs.size[/archive,total]]
44229:20160408:190850.006 listener #1 [waiting for connection]
44228:20160408:190850.114 collector [processing data]
44228:20160408:190850.114 In update_cpustats()
44228:20160408:190850.114 End of update_cpustats()
44228:20160408:190850.115 collector [idle 1 sec]
44231:20160408:190851.048 listener #3 [processing request]
44231:20160408:190851.048 Requested [vfs.fs.size[/archive/ease2n/source,total]]
44231:20160408:190851.049 listener #3 [waiting for connection]
44228:20160408:190851.115 collector [processing data]
44228:20160408:190851.115 In update_cpustats()
44228:20160408:190851.116 End of update_cpustats()
44228:20160408:190851.116 collector [idle 1 sec]
44229:20160408:190852.022 listener #1 [processing request]
44229:20160408:190852.023 Requested [vfs.fs.size[/archive/geographic,total]]
44229:20160408:190852.023 listener #1 [waiting for connection]
44228:20160408:190852.116 collector [processing data]

[kloczek]

I don't see anything suspicious/wrong within you agent logs.

First step to change passive monitoring to active one is change type type of all items from "zabbix agent" to "zabbix (active) agent" first of all templated items and after this untemplated items.

Second step is change agents setup from passive to active one (ServerActive=<proxy.or.server.name> and StartAgents=0 on agent side).