WARNING: Result from SNMP not valid. Partial Result:

[eger]

I searched and could only find a couple posts with these warnings. but were related to other problems. I gett logs filled very quickly with all these warnings and was wondering if there was a fix or any idea why these would be popping up? I am just using the Cisco Router templates to graph lots of switchs. here si what I keep getting lots of in my logs:

Code: Select all

01/01/2005 03:50:02 PM - CMDPHP: Poller[0] Host[2] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:02 PM - CMDPHP: Poller[0] Host[2] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:02 PM - CMDPHP: Poller[0] Host[2] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:02 PM - CMDPHP: Poller[0] Host[2] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:07 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:07 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:07 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:07 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:08 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:09 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:11 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:15 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:17 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:22 PM - CMDPHP: Poller[0] Host[9] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:40 PM - CMDPHP: Poller[0] Host[10] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:50:48 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:05 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:07 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:10 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:12 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:14 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:17 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:19 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:
01/01/2005 03:51:21 PM - CMDPHP: Poller[0] Host[11] WARNING: Result from SNMP not valid.  Partial Result:

I have checked and can't find any odd characters or values being returned by the switches. They are mostly Cisco 2900XL switches and some 3COM.

Thanks!

[Thomas]

I find this messages in my logs as well.
I think this happens when you query an OID that doesn't return any data. I see this problem when I use the Windows Host template against a W2K machine. This kind of problem doesn't occur with a W2K3 maschine.

[eger]

hmm wonder if this is a bug. or if i should be redisgning all my templates to fit that specific switch.

maybe this could be a future option (do not report partial results)? i guess it is a warning though.

[jcaesar]

You can turn the logging level up to debug for a polling cycle and see which OIDs aren't returning valid values. You can use snmpget or some other SNMP tool to see if you are polling for a value that exists.

More often than not, I was trying to graph something that didn't exist on the router (IOS too old to have the correct OIDs, or something similar). Although, once I received these messages when the device was returning valid data and it seems to have been related to the version of PHP I was using. I had to downgrade to a version that I knew worked.

[rony]

I have ran into an interesting problem...

When I was polling the number of users on a AIX 5.2 box, I kept getting the partial results message for that OID. Upon quering the OID manually, NET-SNMP snmpget generated a warning that it returned a "Wrong Type (should be Gauge32 or Unsigned32)", and the value returned was an integer.

As it stands right now, I don't know if this behavor is AIX fault and if cacti should be more forgiving about types returned. I have to look into it and see what I can find.

[TheWitness]

Rony,

You may have conflicting MIB files in Net-SNMP or a bad version of Net-SNMP. Sometimes Net-SNMP get's confused if you have multiple MIB's declaring the same OID. However, I would suspect that Net-SNMP may in fact have a problem.

Larry

[TheWitness]

Another correction would be to delete the AIX mib file. I don't know about forgiveness here, but it's something to consider.

Larry

[rony]

Nope, this is the IBM supplied AIX hostmib agent for the IBM supplies SNMP daemon.

Alternative is to change the mib that the hostmib agent is loading to correct the issue. But my concern is that the result, all be it, incorrect, is still a number. And there should be a way to use it anyways. There are implementations that are wrong that you can't modify the mib or values on.

[brycenesbitt]

Tip: To look up the host numbers "Host[x]" see host.php of the "host" table in mysql.

[rdh043000]

I'm assuming the warning message occurs only when no data is returned. I'm adding 4 data sources for every port of every switch, regardless if a user is plugged into the port or not.

So these lines in my log make it an annoyance to wade through. I went to the poller options and Poller Warnings is unchecked, my log level is set to LOW.

Should I be seeing these warning messages still?

[TheWitness]

The latest version of spine discards messages for ports/devices that have non-working MIBS until you switch to MEDIUM logging level.

TheWitness