Zabbix Proxy Stops Collecting Data - SNMP

[chivasregal]

Hi - I have a zabbix_proxy collecting data from a network device (via snmp) - it all works fine initially, however after about say 30mins-1hr, it just stops collecting data.

No errors or mention of anything in either log (zabbix_proxy log on Proxy or Server Log).

Connectivity to the device is fine, i can ping it and query via snmpwalk. Just Zabbix stops monitoring it?



Server Details:
Zabbix Appliance - Latest (downloaded yesterday)
Proxy - Ubuntu 14.04, Zabbix compiled from the latest source

Note: The Zabbix Server is monitoring alot of other devices/servers (on the local network) and this is working fine. It only seems to be this one proxy which is causing the issue.

The only correlation i have at the moment is when the data stops, i check the proxy log and can see that the housekeeper function has executed. Could this be related?


defsys@CMGNM01:~$ tail -f /var/log/zabbix-proxy/zabbix_proxy.log
1202:20140426:070120.498 proxy #4 started [poller #1]
1206:20140426:070120.499 proxy #8 started [poller #5]
1207:20140426:070120.500 proxy #9 started [unreachable poller #1]
1204:20140426:070120.500 proxy #6 started [poller #3]
1205:20140426:070120.502 proxy #7 started [poller #4]
1203:20140426:070120.502 proxy #5 started [poller #2]
1221:20140426:070120.503 proxy #18 started [discoverer #1]
1219:20140426:080119.776 executing housekeeper
1219:20140426:080119.927 housekeeper [deleted 8609 records in 0.148753 sec, idle 3600 sec]
1199:20140426:080120.355 Received configuration data from server. Datalen 34936

Any ideas?

To note: restarting zabbix_proxy fixes it, but again - for 30mins-1hr and then the same thing.

[chivasregal]

OK - Seems there is some useful logging now. When the data stops collecting, i see the following from the Server (and all the Items in the GUI for that host change status to 'Not Supported')

2053:20140426:090012.013 item [CLNCMPGW01:ifInOctets[PortA]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.013 item [CLNCMPGW01:ifInOctets[PortB]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.013 item [CLNCMPGW01:ifInOctets[PortC]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.013 item [CLNCMPGW01:ifInOctets[PortD]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.013 item [CLNCMPGW01:ifInOctets[PortE]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.014 item [CLNCMPGW01:ifInOctets[PortF]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.014 item [CLNCMPGW01:ifInOctets[PortG]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.014 item [CLNCMPGW01:ifInOctets[PortH]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.014 item [CLNCMPGW01:ifInOctets[sit0]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.014 item [CLNCMPGW01:ifInOctets[imq0]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.
2053:20140426:090012.015 item [CLNCMPGW01:ifInOctets[imq1]] became not supported: SNMP error: (noSuchName) There is no such variable name in this MIB.



Restarting the zabbix_proxy gets it working again, but only for a short-while. Could this be a corrupted install/OS/server issue?

UPDATE:

It seems after 20mins or so, the item status changed again to supported (again for a short-while).

2054:20140426:091012.190 item [CLNCMPGW01:ifOperStatus[PortA]] became supported
2054:20140426:091012.190 item [CLNCMPGW01:ifOperStatus[PortB]] became supported
2054:20140426:091012.190 item [CLNCMPGW01:ifOperStatus[PortC]] became supported
2054:20140426:091012.190 item [CLNCMPGW01:ifOperStatus[PortD]] became supported
2054:20140426:091012.191 item [CLNCMPGW01:ifOperStatus[PortF]] became supported
2054:20140426:091012.191 item [CLNCMPGW01:ifOperStatus[sit0]] became supported
2054:20140426:091012.191 item [CLNCMPGW01:ifOperStatus[imq0]] became supported
2054:20140426:091012.191 item [CLNCMPGW01:ifOperStatus[imq1]] became supported


There is definitely no issue with connectivity to this device (i run other real-time services which utlise the same WAN connection with no issue)

Any ideas would be greatly appreciated.

[dziadzi]

Hi,

I have the same issue afetr upgrading from 2.2.2 to 2.2.3 (same config, same database). I do not use proxy.
No clue where the reason is....

[cminoru]

Same problem

Guys, have the same problem here: some snmp items just stop collecting data.

Front-End messages are: SNMP error: (noSuchName) There is no such variable name in this MIB.

This problem happen in v2.2.3, v2.2.4, v.2.2.5. Tested yesterday on newest v.2.2.6 with no improvements.

Any ideas?

[dziadzi]

Guys, have the same problem here: some snmp items just stop collecting data.

Front-End messages are: SNMP error: (noSuchName) There is no such variable name in this MIB.

This problem happen in v2.2.3, v2.2.4, v.2.2.5. Tested yesterday on newest v.2.2.6 with no improvements.

Any ideas?

I think this is related to SNMP bulk requests introduced in 2.2.3.
Some devices are not fully compliant with SNMP standard and having issues with bulk request.
They provided patch for 2.2.3 and 2.2.4 to disable this but not for furter versions.
However even with this patch seems to me that some devices in my case are not monitored at all. Anyway plese try.
See https://support.zabbix.com/browse/ZBXNEXT-2301
and also my similar issue I raised

https://support.zabbix.com/browse/ZBX-8191

[cminoru]

Thank you man,

This patch just solved my problem!