I am getting OID errors across the board - lots of them. I have Cisco switches, Windows machines (all with SNMP Informant), SuSE boxes and RedHat boxes. They are all reporting oid errors of some sort or another (from what I can see in the logs and the sporadic graphing of all my devices).
I will try removing and rediscovering a couple of devices and see if that solves the problem with those particular devices. >>> On Mon, Nov 13, 2006 at 6:02 PM, in message <[EMAIL PROTECTED]>, Eric Newton <[EMAIL PROTECTED]> wrote: > Hi Dave, > > 175 out of 176 devices is a good sign, however, it looks like some of > your devices are rejecting the OID requests . > > After a configuration cycle, zenperfsnmp will query a device for all the > OIDs that are configured against it. If a device fails to respond to > individual OIDs, or gives an error on that OID, it is removed from the > list of OIDs for that device until the next configuration cycle. > > When the OID is removed, you get that log message. The collects between > config cycles should be fairly quiet. > > Some causes for the failure: > > - snmpinformant is not installed on your windows device > - your net- snmp is not configured on your unix device to send all > the OIDs > - your device doesn't really support the OIDs you are requesting > - SELinux is configured on your machine and net- snmp cannot provide > the data > > I would look for a trend in the collection: what graphs aren't being > collect, and against what devices? > For example, are they are all windows devices? Are they all from the > same brand of router? Are they all from RedHat Linux machines? > > You may want to remove odin and re- add it to see if you can iron out the > situation with it's IP address. > > - Eric > > > DAVE CUSHING wrote: >> Ok, after combing through the logs I did find the timing message >> >> 2006- 11- 13 14:03:06 INFO zen.zenperfsnmp: collected 175 of 176 devices >> in 40.09 >> 2006- 11- 13 14:03:12 ERROR zen.zenperfsnmp: Device odin is reporting >> more than once >> 2006- 11- 13 14:03:12 WARNING zen.zenperfsnmp: snmp agent down on device >> odin >> 2006- 11- 13 14:03:26 INFO zen.zenperfsnmp: Unresponsive devices: >> [['10.0.220.5', 1, 1655]] >> 2006- 11- 13 14:04:15 INFO zen.zenperfsnmp: collected 175 of 176 devices >> in 48.59 >> >> My question is, should I be getting all those oid error messages? I am >> not getting graphs consistently. >> >> If I can supply any other information, or try anything else out, please >> give me a shout. >> >> >> >>>>> On Wed, Nov 8, 2006 at 2:12 PM, in message >>>>> >> <[EMAIL PROTECTED]>, >> Eric Newton <[EMAIL PROTECTED]> wrote: >> >>> Hi Dave, >>> >>> If you could check a few things for me: >>> >>> what is your SNMP cycle time (Monitors- >Performance- >>> localhost)? >>> >>> Make sure that the step and heartbeat agree with this number: >>> >>> $ cd $ZENHOME >>> $ find . - name *.rrd - print0 | xargs - 0 -- max- lines=1 >>> >> rrdtool dump | >> >>> egrep 'step|heartbeat' >>> >>> The heartbeat should be 3x the step. >>> >>> Next, we need to make sure that zenperfsnmp is running through all >>> >> it's >> >>> devices. Run it with debug output: >>> >>> $ zenperfsnmp run - v 10 -- cycle >>> >>> It should stop collecting when it gets to: >>> >>> INFO: collected 303 of 304 devices in 42.15 >>> >>> If you fail to see this message, or get wild numbers like "2 of 100" >>> >> >> >>> then you know something is failing. >>> >>> This "reporting more than once" might happen if devices are modeled >>> under multiple IP addresses, but that's not supposed to happen. We >>> >> can >> >>> check for things like that if the stuff above checks out. >>> >>> - Eric >>> >>> DAVE CUSHING wrote: >>> >>>> But what is causing the error of getting more than one report back >>>> >> from the >> >>> device? I have tried re- installing 0.23 from scratch but the >>> >> problem still >> >>> exists. I have played with zenstep a number of different times, but >>> >> I still >> >>> cannot get the graphing to work without some sort of gaps in it. >>> >>>> Can I retrograde to 0.22 with the current schema structure or will >>>> >> that >> >>> cause problems? >>> >>>> >>>> >>>>>>> On 11/6/2006 at 8:57 PM, in message >>>>>>> >>>>>>> >>>> <304AD09F- 889A- 43D4- 8C50- [EMAIL PROTECTED]>, "Erik A. >>>> >> Dahl" >> >>>> <[EMAIL PROTECTED]> wrote: >>>> >>>> >>>>> Dave, >>>>> >>>>> This is most likely the reason for the other errors you are getting >>>>> >> >> >>>>> (about minimum update times). The "other" Eric is out for a couple >>>>> >> >> >>>>> of days maybe he will have more insight... >>>>> >>>>> - EAD >>>>> >>>>> On Nov 5, 2006, at 12:13 PM, DAVE CUSHING wrote: >>>>> >>>>> >>>>> >>>>>> I don't know if it means anything, but I am also getting a lot of >>>>>> >> >> >>>>>> messages like the one below: >>>>>> >>>>>> 2006- 11- 05 12:07:48 ERROR zen.zenperfsnmp: Device 3406- 1900C- 1 >>>>>> >> is >> >>>>>> reporting more than once >>>>>> 2006- 11- 05 12:07:48 ERROR zen.zenperfsnmp: Device 1708- 3548XL- >>>>>> >> 1 is >> >>>>>> reporting more than once >>>>>> 2006- 11- 05 12:07:49 ERROR zen.zenperfsnmp: Device 2478- 3548XL- >>>>>> >> 1 is >> >>>>>> reporting more than once >>>>>> 2006- 11- 05 12:07:50 ERROR zen.zenperfsnmp: Device >>>>>> Enterprise- 3548XL- 1 is reporting more than once >>>>>> 2006- 11- 05 12:07:50 ERROR zen.zenperfsnmp: Device 1407- 3548XL- >>>>>> >> 1 is >> >>>>>> reporting more than once >>>>>> 2006- 11- 05 12:07:51 ERROR zen.zenperfsnmp: Device 1717A- 3548XL- >>>>>> >> 1 is >> >>>>>> reporting more than once >>>>>> 2006- 11- 05 12:07:52 ERROR zen.zenperfsnmp: Device 4116- 3548XL- >>>>>> >> 1 is >> >>>>>> reporting more than once >>>>>> 2006- 11- 05 12:07:53 ERROR zen.zenperfsnmp: Device 3116- 3548XL- >>>>>> >> 1 is >> >>>>>> reporting more than once >>>>>> >>>>>> I have tried a number of things over the weekend, but none seem to >>>>>> >> >> >>>>>> change the problem I am seeing. Any suggestions would be >>>>>> appreciated, I am stumped at the moment. >>>>>> > > _______________________________________________ > zenoss- users mailing list > zenoss- [EMAIL PROTECTED] > http://lists.zenoss.org/mailman/listinfo/zenoss- users _______________________________________________ zenoss-users mailing list [email protected] http://lists.zenoss.org/mailman/listinfo/zenoss-users
