No active checks on server: host [] not found

[mobileum]

Good morning everybody, i have a problem with a few servers im trying to monitor and its driving me crazy. I added 5 new servers today, all within a private network (10.100.100.6 ... 10.100.100.11), and only one's working. All those servers have the same config, all active checks, and still, 5 of them give me the error "no active checks on server: host [] not found". I've checked 100 times that the hostname in the zabbix frontend matches the hostname in the agentd.conf, and tried several hostnames, all of different lengths, with/without numbers, etc..., and neither worked. When i turn up the debuglevel up to 4, all i see on the server's log is:

 22001:20150107:162637.478 Starting Zabbix Agent [lebfra]. Zabbix 2.0.12 (revision 45390).
 22001:20150107:162637.479 In init_collector_data()
 22001:20150107:162637.480 End of init_collector_data()
 22002:20150107:162637.482 agent #0 started [collector]
 22002:20150107:162637.483 In init_cpu_collector()
 22003:20150107:162637.483 agent #1 started [active checks]
 22003:20150107:162637.484 In init_active_metrics()
 22003:20150107:162637.484 Buffer: first allocation for 100 elements
 22002:20150107:162637.485 End of init_cpu_collector():SUCCEED
 22002:20150107:162637.486 In update_cpustats()
 22003:20150107:162637.484 In send_buffer() host:'200.45.235.164' port:10051 values:0/100
 22003:20150107:162637.488 End of send_buffer():SUCCEED
 22003:20150107:162637.488 refresh_active_checks() host:'200.45.235.164' port:10051
 22002:20150107:162637.489 End of update_cpustats()
 22003:20150107:162637.797 sending [{
        "request":"active checks",
        "host":"lebfra"}]
 22003:20150107:162637.798 before read
 22003:20150107:162638.428 got [{
        "response":"failed",
        "info":"host [lebfra] not found"}]
 22003:20150107:162638.428 In parse_list_of_checks()
 22003:20150107:162638.428 In disable_all_metrics()
 22003:20150107:162638.429 no active checks on server: host [lebfra] not found
 22003:20150107:162638.429 In process_active_checks('200.45.235.164',10051)
 22003:20150107:162638.429 End of process_active_checks()
 22003:20150107:162638.430 In get_min_nextcheck()
 22003:20150107:162638.430 In send_buffer() host:'200.45.235.164' port:10051 values:0/100
 22003:20150107:162638.430 End of send_buffer():SUCCEED
 22003:20150107:162638.431 Sleeping for 1 second(s)

Same message in the zabbix server's side

28985:20150107:164446.988 cannot send list of active checks to [10.16.3.195]: host [lebfra] not found

I've checked that the zabbix server port 10051 is accesible from the servers with telnet, and that they both can connect succesfully. I've even tried to add the two of them in a specific vpn, and that didin't help, so connectivity issues are ruled out.

Does anybody have any idea on how to proceed? Any help would be deeply appreciated.

Thanks!!

[tchjts1]

I've checked 100 times that the hostname in the zabbix frontend matches the hostname in the agentd.conf

And also in zabbix_agentd.conf, you checked that the field ServerActive= is the IP or DNS of your Zabbix server?

I have also seen cases where people are modifying the wrong file because they have multiple copies on their hosts.

[mobileum]

Hi, thank you very much for your answer. Im only editing the file in /etc/zabbix/agentd.conf, the other files that i have are

/root/zabbix_agentd.conf
/root/zabbix-2.0.12/conf/zabbix_agentd.conf
/usr/local/etc/zabbix_agentd.conf
/usr/local/etc/zabbix_agentd.conf.d

and i dont think those are the ones i need to edit. As with the serveractive parameter, yes, i have the same ip as in the only working host in this segment (and 100+ hosts in other segments). I've run out of ideas..

[tchjts1]

Certainly you can specify whichever conf file you wish to use if you are using the -c /etc/zabbix/zabbix_agentd.conf switch and path. But I believe by default it is the one I have in bold below. I may be wrong. I don't recall in which version they changed the default location.

/root/zabbix_agentd.conf
/root/zabbix-2.0.12/conf/zabbix_agentd.conf
/usr/local/etc/zabbix_agentd.conf
/usr/local/etc/zabbix_agentd.conf.d

[mobileum]

I've just checked that an according to zabbix documentation, the one in /etc/zabbix/zabbix_agentd.conf was correct

"-c --config <file> specify configuration file, default is /etc/zabbix/zabbix_agentd.conf"

[tchjts1]

Are these Linux hosts? What are your SELinux settings?

You can run sestatus to see what it is.

[mobileum]

Hi!, these are all linux hosts, i don't have that packet installed (policycoreutils), so i can't give you the output of sestatus, but i really don't think thats the problem, i don't have it installed in any of the other servers monitored by zabbix, nor in the other server thats working with the exact configuration in the same segment. All these servers have a very basic installation of ubuntu 12 server edition.

[gleepwurp]

cannot send list of active checks to [10.16.3.195]: host [lebfra] not found

Well, communication seems ok, as you see from this message that the agent was requesting a list of ITEMS to download, but the Zabbix server can't find the host object in its database.

Does your hostname "case" match on your zabbix server? ie: are they both lowercase?

from the error message I see, looks like the Zabbix server doesn't _have_ a host object defined under the name lebfra... Either the hostname case is wrong, or you don't have the host defined?

Also, the hostname match is done on the "host name" field, and not the "visible name" in the host configuration screen.

G.

[tchjts1]

gleepwurp has a good point. If I run into a similar issue on my setup and suspect there is an issue with the name, I simply copy and paste whatever is in zabbix_agentd.conf directly into the frontend to rule out any misspellings.

And honestly, I haven't put a name into Hostname= ever since auto-registration was introduced.

[mobileum]

Hi, yes, i've checked the hostname, tried to copy and paste it, checked the lowercase and even deleted and recreated the host in zabbix multiple times, all without results. Again, i have another host thats working perfectly, but just that one, everythings similar. Is there any other issue that could be causing this error message?

[gleepwurp]

Really weird...

I would try to export the host from the Zabbix server into an xml file and open that file and see if there might be anything amiss...?

You probably tried this one already, but just in case: Delete the Host from Zabbix server and recreate it.

If you want, you could attach your xml export to your next message (and edit out any sensitive information before you do), and I'll take a peek at it to see if anything jumps out. 2 pairs of eyes are better than one!

G.