Stop Collect data

[maruscya]

Hello guys.

I have a strange problem on server. I try explain the problem. I buy a new server for zabbix service, Xeon Cpu (quad core) with 5Gb and raid 1 system

Ok, gentoo works realy fine, and zabbix (compiled from source) too. At the moment there are 30 host with around 10 items per host (item are setup as active). Ok i think that my hardware can manage more than 30 host

All agents report this message into log file

19716:20080208:102846 Timeout while answering request
19716:20080208:102846 Getting list of active checks failed. Will retry after 60 seconds

it's not a network problem (firewall, router etc..) message appear on all macchines , include server in the same lan. I'm using the standard configuration for zabbix server (6 poller and 5 trapper) and default configuration for agent (StartAgent=5)

If i check some graph (like cpu load etc) and many data is lost..

This graph report a number of message in queue.... i have setup a trigger for this item, and trigger never fired up becuse zabbix loose a lot of data.

EDIT: netstat command report more than 3000 connection on zabbix port ... Can be a kernel problem ? Like too many connection on tcp port??


Any suggestion ???

I hope that some problems can be solved/fixed in riga next mouth

[maruscya]

Update..

I try install nightly release of zabbix.... i try connect from server to zabbix port, but .... the connection is reeeeeealy slow...

I think there are some problems on zabbix server... but i don' t understand .... where is...

[cbidwell]

Are your client(s) on the same physical subnet as the server? Perhaps it's network latency? Did you confirm firewall(s) have required policies allowing tcp/10050 to allow through?

[maruscya]

Are your client(s) on the same physical subnet as the server? Perhaps it's network latency? Did you confirm firewall(s) have required policies allowing tcp/10050 to allow through?

I made some tests, is not a network problem. between server and client there aren't any limitation (firewall etc..) .

I try to rebuild DB, and change many setting : table engine, allocation memory etc...

At the moment server seems works... but... is again in a "test mode"...

I hope that alexei, will explain me some tricks about configuration during next training ...

[maruscya]

Solved - I hope!!

Hello Guys..

I have fixed this issue changing timeout into zabbix_agentd.conf.

Timeout = 3

now is

Timeout = 6

Restart agent and it seems works fine