Problem pulling a host - Recache DataQuery

[fedeglobant]

I'm running cacti 0.8.7e with spine. Trying to monitor a new linux machine I got this in the logs

01/29/2013 02:25:23 PM - SPINE: Poller[0] Host[282] TH[1] Recache DataQuery[1] OID: .1.3.6.1.2.1.1.3.0, output: U
01/29/2013 02:25:23 PM - SPINE: Poller[0] Host[282] TH[1] Recache DataQuery[2] OID: .1.3.6.1.2.1.1.3.0, output: U
01/29/2013 02:25:23 PM - SPINE: Poller[0] Host[282] TH[1] NOTE: There are '2' Polling Items for this Host
01/29/2013 02:25:23 PM - SPINE: Poller[0] Host[282] TH[1] Total Time: 2 Seconds

And no matter what I do I can not get it to pull data from it. I have disabled the availabillity check, deleted the device and added it again, poller_reindex_hosts.php, rebuild_poller_cache.php and nothing worked.

Does anyone have any ideas?

Thanks in advance.

[BSOD2600]

does basic snmp queries against the device work?

[fedeglobant]

Yes it does.

I'm using SNMP v3, don't know if it makes any difference.

[BSOD2600]

does cacti consider Host[282] up?
snmp error shown for it in devices?
You're able to snmpwalk .1.3.6.1.2.1.1.3.0 from the cacti server to Host[282]?

[fedeglobant]

does cacti consider Host[282] up?
I have availability checks disabled so it is up.

snmp error shown for it in devices?
No errors

You're able to snmpwalk .1.3.6.1.2.1.1.3.0 from the cacti server to Host[282]?
Yes I can

What I think it's happening is that there were errors at some point (some kind of connectivity problem) and now it's fine. But for some reason Cacti still sees this device as Down or something. I tryed deleting the device and creating it again and the logs are exactly the same, like there is some kind of cache regarding the ip address of that device. Tell me if I'm taking nonsense...

Thanks again for your help.

[fedeglobant]

Another thing, if I delete the host and create it again the SNMP cache remains, I don't know why.

[BSOD2600]

But for some reason Cacti still sees this device as Down or something. I tryed deleting the device and creating it again and the logs are exactly the same, like there is some kind of cache regarding the ip address of that device. Tell me if I'm taking nonsense...

1) are there any data sources for cacti to poll against that device? IF not, cacti won't check if it's up or not...
2) rebuild the poller cache.

[fedeglobant]

I erased the host, rebuild cache, reindex hosts, rebuild cache again, added the host again and now I get this...

02/06/2013 07:15:25 PM - SPINE: Poller[0] Host[286] SNMP Ping Error: Unknown error: 2
02/06/2013 07:15:25 PM - SPINE: Poller[0] Host[286] TH[1] NOTE: There are '2' Polling Items for this Host
02/06/2013 07:15:25 PM - SPINE: Poller[0] Host[286] TH[1] Total Time: 4.1 Seconds

I have no idea what's going on...