Skip navigation
1 2 Previous Next 23545 Views 15 Replies Latest reply: Jun 23, 2010 11:40 AM by nilie RSS
peak6 Rank: White Belt 51 posts since
Feb 25, 2010
Currently Being Moderated

Mar 18, 2010 4:17 PM

Cisco ethernetCsmacd_64 OID bad - yet good

So When I model my Cisco device I get errors stating

Error reading value for "ifHCInOctets" on '***** (oid .1.3.6.1.2.1.31.1.1.1.6 is bad)
Error reading value for "ifHCInUcastPkts" on '*****(oid .1.3.6.1.2.1.31.1.1.1.7 is bad)
Error reading value for "ifHCOutOctets" on '***** (oid .1.3.6.1.2.1.31.1.1.1.10 is bad)
Error reading value for "ifHCOutUcastPkts" on '***** (oid .1.3.6.1.2.1.31.1.1.1.11 is bad)
Error reading value for "ifInErrors" on '***** (oid .1.3.6.1.2.1.2.2.1.14 is bad)
Error reading value for "ifOutErrors" on '***** (oid .1.3.6.1.2.1.2.2.1.20 is bad)

When I go to the EthernetCSMACD_64 and run a test against the same device inside the template I get results back that look fine:

 

_____

IF-MIB::ifHCInOctets.143 = Counter64: 43931571738 IF-MIB::ifHCInOctets.144 = Counter64: 13871064782 IF-MIB::ifHCInOctets.145 = Counter64: 11811305408 IF-MIB::ifHCInOctets.146 = Counter64: 2250349032 IF-MIB::ifHCInOctets.147 = Counter64: 8494714789 IF-MIB::ifHCInOctets.148 = Counter64: 290545471 IF-MIB::ifHCInOctets.149 = Counter64: 3615952 IF-MIB::ifHCInOctets.150 = Counter64: 6453664651 IF-MIB::ifHCInOctets.151 = Counter64: 25752391332 IF-MIB::ifHCInOctets.152 = Counter64: 29246752508 IF-MIB::ifHCInOctets.153 = Counter64: 105093985224 IF-MIB::ifHCInOctets.154 = Counter64: 74715702429 IF-MIB::ifHCInOctets.155 = Counter64: 18123651654 IF-MIB::ifHCInOctets.156 = Counter64: 23455456141 IF-MIB::ifHCInOctets.157 = Counter64: 12588365809 IF-MIB::ifHCInOctets.158 = Counter64: 25594463923

____

more data of course, but thats a snippet.

I can walk it fine from the server as well. So the OID is correct, Zenoss recognizes it ok when done manually through the template, but it throws errors. I disabled all non active ports, and removed vlans just incase it was trying to read those. Whenever I see the bad OID in the forums they say create a new one and remove the 'broken' oids, since its all of them it sorta makes the template useless doing that.. and besides that it works fine if I do it manually....

 

suggestions?

  • guyverix ZenossMaster 846 posts since
    Jul 10, 2007
    Currently Being Moderated
    1. Mar 18, 2010 10:29 PM (in response to peak6)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    Any chance you could turn on debug logging for zenmodel, and remodel the device?  ( I dont know that it will give any extra information, but it might)

     

    Also if this is for a specific interface, take a look at the index number that it is using for the interface.  Perhaps that index changed.  Run the walk by hand and see if that is adding to your problem..

  • jhinge Rank: White Belt 30 posts since
    Mar 17, 2010
    Currently Being Moderated
    4. Mar 26, 2010 8:19 AM (in response to peak6)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    I seem to be seeing them too :

     

    firstTime2010/03/26 08:29:40.000
    eventClass/Perf/Snmp
    messageError reading value for  "GigabitEthernet3/13.2" on cd-sk01-cor01 (oid .1.3.6.1.2.1.2.2.1.20.139  is bad)

     

    the template used to gather data is : ethernetCsmacd if this is helpful ?

  • guyverix ZenossMaster 846 posts since
    Jul 10, 2007
    Currently Being Moderated
    5. Mar 26, 2010 8:35 AM (in response to peak6)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    carpej wrote:

     

    Here is a debug:


    I think ithis might be related

    The threshold high utilization in template /zport/dmd/Devices/rrdTemplates/ethernetCsmacd_64 has caused an exception. (On Zenoss Server)

     

    Has anyone seen this before? I am not sure if its causing an exception because of the OID being bad, or if its unrelated.

     

    This might be the cause.  Disable the threshold in the template, and see what happens.

  • guyverix ZenossMaster 846 posts since
    Jul 10, 2007
    Currently Being Moderated
    6. Mar 26, 2010 8:37 AM (in response to jhinge)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    jhinge wrote:

     

    I seem to be seeing them too :

     

    firstTime2010/03/26 08:29:40.000
    eventClass/Perf/Snmp
    messageError reading value for  "GigabitEthernet3/13.2" on cd-sk01-cor01 (oid .1.3.6.1.2.1.2.2.1.20.139  is bad)

     

    the template used to gather data is : ethernetCsmacd if this is helpful ?

    In a way.  At least we know that it happens on 32 bit and 64 bit counters.

    On a gig link, I am suprised you are using 32 bit due to the counter rollover problems..

     

    are you able to walk the oid by hand and see if it really is bad?

  • jhinge Rank: White Belt 30 posts since
    Mar 17, 2010
    Currently Being Moderated
    7. Mar 26, 2010 9:23 AM (in response to guyverix)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    It is a renamed 64bit counter

     

    snmpwalk gives the following :

    [drift@zorglub ~]$ snmpwalk -v 2c -c xxxxxx 192.168.0.61 sysDescr
    SNMPv2-MIB::sysDescr.0 = STRING: Cisco Internetwork Operating System Software
    IOS (tm) s72033_rp Software (s72033_rp-ADVIPSERVICESK9_WAN-M), Version 12.2(18)SXF17, RELEASE SOFTWARE (fc1)
    Technical Support: http://www.cisco.com/techsupport
    Copyright (c) 1986-2009 by cisco Systems, Inc

     

    [drift@zorglub ~]$ snmpwalk -v 2c -c xxxxxx 192.168.0.61 .1.3.6.1.2.1.2.2.1.20.139
    IF-MIB::ifOutErrors.139 = No Such Instance currently exists at this OID
    [drift@zorglub ~]$ snmpwalk -v 2c -c xxxxxx 192.168.0.61 1.3.6.1.2.1.2.2.1.20.139
    IF-MIB::ifOutErrors.139 = No Such Instance currently exists at this OID
    [drift@zorglub ~]$

     

    quite possibly the above can translate to "I am a 'behind-simian', and the person configuring this should be flogged" but that will also

    teach me something

  • jhinge Rank: White Belt 30 posts since
    Mar 17, 2010
    Currently Being Moderated
    8. Apr 8, 2010 3:06 AM (in response to jhinge)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    I must admit to still seeing this and feeling stumped,

    I'm green enough in this, that I don't quite know what to check/modify to make this go away, except for manually deleting all the pretty gray events from my event console, which is.. like wetting myself to get warm..

  • guyverix ZenossMaster 846 posts since
    Jul 10, 2007
    Currently Being Moderated
    9. Apr 8, 2010 3:51 AM (in response to jhinge)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    I am betting that the device has re-indexed it's OID values..  Try just a walk of the table, not the index, and see if what you are expecting at index 139 is hiding somewhere else..

     

    Grey = bad for a device in my book.. Means it was not put into the correct class, or the class has incorrect definitions for what is being modeled..  Too many of those grey alarms will also affect Zenoss performance..

  • jenkinskj Rank: Green Belt 330 posts since
    Jul 30, 2009
    Currently Being Moderated
    10. Jun 15, 2010 1:59 PM (in response to guyverix)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    Is there any update to this issue? I have the same problem. The interface looks to be monitored but I continue to get errors regarding a bad OID. The OID which lists is one level further down the MIB tree than what is defined in the ethernetCsmacd templates.

     

    I am using SNMPver v2c. I can snmpwalk the device and grep for the OID. For some reason all of the interfaces can not get to this specific OID though.

     

    Can this be suppressed? Is there a work around?

     

    - Ken

  • istoptofly Rank: Green Belt 305 posts since
    Apr 20, 2010
    Currently Being Moderated
    11. Jun 15, 2010 2:15 PM (in response to jenkinskj)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good
  • jenkinskj Rank: Green Belt 330 posts since
    Jul 30, 2009
    Currently Being Moderated
    12. Jun 15, 2010 2:46 PM (in response to istoptofly)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    Thanks John.

     

    I added the transform statement under the event class /Perf/Snmp. I will see if it works.

     

    Do by chance know why this is failing?

     

    - Ken

  • istoptofly Rank: Green Belt 305 posts since
    Apr 20, 2010
    Currently Being Moderated
    13. Jun 15, 2010 2:49 PM (in response to jenkinskj)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    What I've read in the past is the Cisco devices sometimes change their OID's out of the blue, I don't have anymore info than that.

     

    John

  • jenkinskj Rank: Green Belt 330 posts since
    Jul 30, 2009
    Currently Being Moderated
    14. Jun 15, 2010 2:52 PM (in response to istoptofly)
    Re: Cisco ethernetCsmacd_64 OID bad - yet good

    That is good information. I will add it to my event transform description. Thank you for all your help on this one.

     

    - Ken

1 2 Previous Next

More Like This

  • Retrieving data ...

Legend

  • Correct Answers - 4 points
  • Helpful Answers - 2 points