1.6.2. Data from proxy are lag more 1 hour

[kc5kse]

Well, I spoke to soon. Data is already an hour behind once more.

[stemasie]

hmmm strange, but i think you are on the right way...
did you completly re-created the DB? maybe take a quick change to sqlite and see if this is working.
maybe you think about an upgrade to a newer version ...

edit: do you use the same timeserver? you have EXACTLY the same time on proxy and server?

[kc5kse]

I'm using the latest v 1.6.6. My proxies are pretty heavliy used and I think mysql is better suited. We do have some issues with ntp that we are working on but not more than a minute or minute and a half different.

[stemasie]

is it the same effect when you deactive some items or clients?
maybe they're overloaded...

[justincase]

Hello. I had the same problem and I did a couple of things which helped.

- increase the number of pollers in the server config

- change the polling frequency for the network interfaces from 5 seconds up to 30 seconds. I did this in the linux template so all hosts were immediately impacted.

- fixed the hostnames to remove the remove the errors that looked like:
Send list of active checks to [192.168.24.234] failed: host [host.at.mydomain.com] not found

See http://www.zabbix.com/forum/showthre...+active+checks

- Fixed or disabled check that cause these message:
[host.at.mydomain.com:net.if.in[eth0,bytes]] error: Type of received value [1598548278930] is not suitable for value type [Numeric (float)]

See http://www.zabbix.com/forum/showthread.php?t=13588

For some of this I went to Administration->Queue and looked at the details to figure out which hosts specifically were laggy. After making all these changes things are better. I am running 1.6.6 but moving to that did not fix these.

I hope this helps.

/justin

[Alexei]

I think that slow performance of Zabbix Server is the core reason of these problems. Please check that number of trappers is at 2 x <number of proxies>, also check that Zabbix Server has DB cache turned ON (parameter StartDBSyncers=1).

[kc5kse]

I edited zabbix_server.conf and restarted. We'll see how it goes. One thing though; I am unable to view Details from the Queues. I don't know if it's just too much data or what.

[kc5kse]

Ok, now i'm getting "History buffer is full. Sleeping for 1 second." No data is coming across at all.

[danpilla]

Hello. I had the same problem and I did a couple of things which helped.

- increase the number of pollers in the server config

- change the polling frequency for the network interfaces from 5 seconds up to 30 seconds. I did this in the linux template so all hosts were immediately impacted.

- fixed the hostnames to remove the remove the errors that looked like:
Send list of active checks to [192.168.24.234] failed: host [host.at.mydomain.com] not found

See http://www.zabbix.com/forum/showthre...+active+checks

- Fixed or disabled check that cause these message:
[host.at.mydomain.com:net.if.in[eth0,bytes]] error: Type of received value [1598548278930] is not suitable for value type [Numeric (float)]

See http://www.zabbix.com/forum/showthread.php?t=13588

For some of this I went to Administration->Queue and looked at the details to figure out which hosts specifically were laggy. After making all these changes things are better. I am running 1.6.6 but moving to that did not fix these.

I hope this helps.

/justin

I have the same problem, i make that you told, but this didn't resolve my problem, anybody can helpe me with this error?

19866:20091014:133458 Send list of active checks to [192.168.0.70] failed: host [ora02] not found

[kc5kse]

Bump? I've reduced the load on my Master server a good bit, but I'm still seeing an hour or more from proxies. Also, hosts I have disabled are still showing up in the log as "unable to connect...".

Any help would be greatly appreciated!!