Momentary loss of access to most agents

[shaun666]

Hi,

We are running Zabbix 2.0.6 on CentOS 5.10.

I'd like to start off by thanking the Zabbix team for making this great software freely available. It is without doubt the most versatile monitoring system available.

However as of 0104 Hrs this morning our server reports "Zabbix Agent not accessible" for almost all our clients. This problem persists for for about 20 minutes at 4 minutes past every hour, then normal service is resumed.

I've triple checked our network and can confirm the server can reach TCP port 10050 on all the agents every time this issue is reported.

I notice entries like this appear in the logs at 4 minutes passed the hour:

26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:18683 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:22625 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:22765 value:0(0) new_value:0
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:23658 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:30126 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:32417 value:0(1) new_value:2
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:32597 value:0(1) new_value:2
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:34940 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:40797 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:40800 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:40809 value:0(0) new_value:0
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL
26034:20150601:170400.014 In DBget_trigger_update_sql() triggerid:40868 value:0(1) new_value:2
26034:20150601:170400.014 End of DBget_trigger_update_sql():FAIL

I would be grateful for any advice.

TIA

Shaun

[Atsushi]

Probably load of processing that is periodically run it will think it's high.
Zabbix server processing at a set time interval the default is to run.
This is the process for deleting old data.

Setting is the HousekeepingFrequency of the setting in the zabbix_server.conf.
Default value is 1. This is to run every hour.

https://www.zabbix.com/documentation/2.0/manual/appendix/config/zabbix_server

Might database tuning effectively be in lowering the load.
There is also the case to some extent decreases the load on the database
by Zabbix version up.