12-07-2022 06:26 AM
I have a monitoring Infrastructure that regularly polls all of my leaves and spines using SNMP, all of these nodes are a part of the same POD, which is the object on which the SNMP policy is applied.
Some leaves return my requested OIDs, and Some Don't, I can't seem to find any causes for the issue as the same SNMP policy is applied on all nodes.
It is important to note that when the the leaves are polled for all of the OIDs they hold in their MIBs all of the leaves return something. The problem is that for some the data that is returned is extremely minimal.
The polling on the spines work fine
Solved! Go to Solution.
12-07-2022 08:35 PM
Hello,
Please share the output of "show snmp summary" and check if the snmp process is running on the leaf with an OID. Some troubleshooting steps are documented here - https://community.cisco.com/t5/data-center-and-cloud-knowledge-base/technote-snmp-in-the-aci-fabric/ta-p/3157795
You need to check Fabric policy, pod policy and if snmp policy is consumed in them. If you get an unexpected output in the "show snmp summary" then probably there is a missing piece of the configuration. You could share the output of the verification commands listed in the link above to understand the issue better.
Cheers!
12-11-2022 12:49 AM
Never Mind I found the answer, I need to use to community Policies. One for the global scope and one for the VRF context.
@anirukasthank you for all the help
12-07-2022 09:45 AM
After looking again I did find one commonality between all of the leaves that don't return the requested OIDs, they all have down links to servers both connected and conigured. The two leaves that do return all of the OIDs have no down links in up state nor are there any policies applied to the leaves' down links.
Does anyone have any idea why this could be happening?
12-07-2022 08:35 PM
Hello,
Please share the output of "show snmp summary" and check if the snmp process is running on the leaf with an OID. Some troubleshooting steps are documented here - https://community.cisco.com/t5/data-center-and-cloud-knowledge-base/technote-snmp-in-the-aci-fabric/ta-p/3157795
You need to check Fabric policy, pod policy and if snmp policy is consumed in them. If you get an unexpected output in the "show snmp summary" then probably there is a missing piece of the configuration. You could share the output of the verification commands listed in the link above to understand the issue better.
Cheers!
12-11-2022 12:47 AM
Thanks @anirukas, I believe I found the problem but I can't find a solution online,
leaf-1# show snmp summary
Admin State : enabled, running (pid:17310)
Local SNMP engineID: [Hex] 800000090370DF2F0F2B16
[Dec] 128:000:000:009:003:112:223:047:015:043:022
----------------------------------------------------------------------
Community Context Status
----------------------------------------------------------------------
public ok
----------------------------------------------------------------------
User Authentication Privacy Status
----------------------------------------------------------------------
----------------------------------------------------------------------
Context VRF Status
----------------------------------------------------------------------
----------------------------------------------------------------------
Client VRF Status
----------------------------------------------------------------------
192.168.10.0/24 management ok
-------------------------------------------------------------------------------
Host Port Ver Level SecName VRF
-------------------------------------------------------------------------------
leaf-2# show snmp summary
Admin State : enabled, running (pid:17310)
Local SNMP engineID: [Hex] 800000090370DF2F0F2B16
[Dec] 128:000:000:009:003:112:223:047:015:043:022
----------------------------------------------------------------------
Community Context Status
----------------------------------------------------------------------
public test-env ok
----------------------------------------------------------------------
User Authentication Privacy Status
----------------------------------------------------------------------
----------------------------------------------------------------------
Context VRF Status
----------------------------------------------------------------------
test-env test-env:test-env ok
----------------------------------------------------------------------
Client VRF Status
----------------------------------------------------------------------
192.168.10.0/24 management ok
-------------------------------------------------------------------------------
Host Port Ver Level SecName VRF
-------------------------------------------------------------------------------
Leaf-1 return OIDs from the Global Scope while Leaf-2 only returns OIDs associated with VRF context. How can I also request Global Scope OIDs from leaves that are associated with the VRF's snmp context? Sorry if I am being confusing I am not very knowledgeable on the subject of snmp in ACI.
12-11-2022 12:49 AM
Never Mind I found the answer, I need to use to community Policies. One for the global scope and one for the VRF context.
@anirukasthank you for all the help
12-11-2022 11:41 PM
Thats great!, yes it needs community strings in global scope. You also have APIC as SNMP aggregator feature if you want to tryout - https://www.cisco.com/c/en/us/td/docs/dcn/aci/apic/5x/basic-configuration/cisco-apic-basic-configuration-guide-51x/m_management.html#id_61644
Happy ACIng!
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide