No updates

[KBeideman]

Greetings and salutations...

I installed zabbix 1.1.3 on a Ubuntu 5 box. no probs. it monitors about 25 winboxes and about 5 linux boxes. the app worked like a charm for 5 days, then it just suddenly stopped. The data does not get updated. Nothing changed on the box. The server is running and the port appears to be open. Not even the agent on the localbox works. The odd thing: if I change the template for a box then change it back, it will grab updated info....but it will not continue to update from there. the conf files havent changed either. I am stumped. anybody got any ideas?

Thanks
Kirk

[Alexei]

Perhaps system time has been changed to past?

[KBeideman]

Thanks Alexei for your quick response.

Good point that I didnt think of. So I checked the time and it is still synced with the clients. so that didnt change.

I have noticed that it works intermttently. After I change the template it will grab data for about 2-3 minutes then stop. help!

[KBeideman]

Ok... this appears to be a queue related issue. I have 2 boxes that I dont have the client installed on yet i had listed as monitored in Zabbix. The queue ended up having over a thousand in the 5 minutes + column. Need a way to flush the queue.

[dantheman]

There really isn't a way to "flush" the queue, but you can disable the hosts that do not yet have the agent installed, that would remove those items from the queue and should get things working again.

[KBeideman]

This is very frustrating....

As soon as zabbix_server cant contact a host the queue starts to backup. And this isnt just because I havent installed all the clients. I stopped monitoring all hosts that dont have the client. Some of our people turn off their computers at night, or some laptops travel...

This is what happens...

1. zabbix_server cant contact host (pc turned off)
2. zabbix queue gets overloaded
3. zabbix_server crashes
4. restart zabbix_server
5. zabbix_server starts parsing the queue, but nothing else gets updated!
6. the only way I've been able to restart updating is to unassign then reassign a template to the host.
7. doing this deletes all old data! therefore the only history I have on the hosts is less than 24 hours!
8. help

[[email protected]]

Did you ever get any information relating to this that could help you out? I have the same problem with my queue - it's becoming too manually intensive to keep active monitoring up on devices which dont receive updates.

[Alexei]

I would suggest tuning of timout relates configuration parameter of ZABBIX Server. Please see ZABBIX Manual (PDF) for more details about these parameters.