- To allow SNMP communications, you must configure an ‘out-of-band contract’ in ‘mgmt’ tenant allowing SNMP traffic. SNMP traffic typically uses UDP port 161 for SNMP requests.
- Configure the APIC out-of-band IP addresses in the ‘mgmt’ tenant. Although the out-of-band addresses are configured during APIC setup, the addresses must be explicitly configured in the ‘mgmt’ tenant before the out-of-band contract will take effect.
ACI provides SNMPv1, v2, and v3 support, including Management Information Bases (MIBs) and notifications (traps). The SNMP standard allows any third-party applications that support the different MIBs to manage and monitor the ACI fabric.
– SNMP read queries (Get, Next, Bulk, Walk) are supported by leaf, spine switches and by APIC.
– SNMP write commands (Set) are not supported by leaf, spine switches or by APIC.
– SNMP traps (v1, v2c, and v3) are supported by leaf and spine switches and by APIC.
– SNMPv3 is supported by leaf and spine switches and by APIC.
Note: Maximum of 10 trap receivers are supported. If you configure more than 10, some will not receive notifications. (Servers listen in on port 162. ACI agents listen in on port 161)
For the complete list of MIBs supported in ACI, see http://www.cisco.com/c/en/us/td/docs/switches/datacenter/aci/apic/sw/1-x/mib/list/mib-support.html
Allowing SNMP access to the APIC Controllers
Unlike Fabric switches, a contract is required in order for the APIC to allow SNMP. For SNMP to successfully contact your APICs, you will need configure a couple of items:
- You will have to ensure that your OOB contract permits SNMP (Ports 161/162)
- You will have to make sure that you have Node Management Addresses in Tenant Mgmt configured for all of your APICs.
Step 1 – Management Contracts (Out-of-band and Inband)
If you already have OOB or Inband Contracts defined inside of Tenant Mgmt, you will need to add the appropriate filters to your contracts for SNMP (UDP-161 and UDP-162). NOTE – If you are using a permit any contract, then you can skip this step)
- Add UDP-161 to existing Contract (if you are not using default/common or permit any)
- Tenants > Tenant mgmt > Security Policies
- Expand Out-of-Band Contracts
- Edit existing OOB Contract
- Select OOB Subject
- Select OOB Filter
- Review filter and add (if necessary) UDP-161
- If you also have Inband Mgmt Connectivity configured, also verify the INB contract filter to permit UDP-161.
- Verify you are providing Out-of-Band Contract in Tenant mgmt
- Tenant > Tenant mgmt > Node Management EPGs > Out-of-Band EPG default
- Under the “Provided Out-of-Band Contracts” in the policy window, provide the appropriate contract (this could be a the default/common contract, or a specific contract you have created and modified).
- Verify you are consuming Out-of-Band Contract in Tenant mgmt
- Tenant > Tenant mgmt > External Management Network Instance Profiles > YourInstanceProfile
- Consume the same contract which you provided in the previous step
- Enter the subnets which are allowed to have access to the APIC (0.0.0.0/0 will permit all)
Step 2 – Configure (or check) to ensure APICs are configured in the Node Management Addresses
Tenant > Tenant mgmt > Node Management Addresses > Static Node Management Addresses
- Configure a separate entry for each of your APICs
- Node IDs for your APIC will range from 1-3 (assuming you have a 3-node APIC cluster).
Configuring SNMP for your Fabric Switches
There are two different places where you need to configure SNMP for your ACI Fabric Switches in the GUI:
- Under Admin Tab
- Under Fabric Tab
Under Admin Tab
Trap Destinations are configured here.
Goto Admin > External Data Collectors > Monitoring Destinations > SNMP > [right click] Create SNMP Monitoring Destination Group
Step 1. Profile
Give it a name (e.g. SNMP-Traps-Receivers) then Click Next
STEP 2 > Trap Destinations
Click the ‘+’ sign to add NMS Servers (receivers)
Then add the destination servers. The required parameters here are Server IP (or hostname if DNS is configured), Community password, and Management EPG. In our case we have them as below:
Host IP: 10.1.2.3
Community Name: public
Management EPG: default (Out-of-Band) <<<<<<<<<<< this out-of-band management was configured separately and is a pre-requisite for SNMP to work.
Then click Finish.
Repeat the above process to configure more destinations for SNMP traps.
This has completed the task of configuring NMS servers to receive traps. We will use this server IP (or a set of servers) as destinations in other parts of the configuration as well.
Under Fabric Tab
Under the Fabric menu, we will configure at two places:
- Pod Policies
- Monitoring Policies
- Pod Policies
Goto Fabric > Fabric Policies > Pod Policies > SNMP >[right click] Create SNMP Policy. You will see the below screen and fill it according to the sample shown below.
Name – give it a name
Admin State – Enable
Contact – give a contact name here (optional)
Location – give the location of the DC (optional)
Community Policies – click the “+” sign and enter the community password here. Then hit Update
Then click the “+” sign under “Client Group Policies”
You will enter into a window “Create SNMP Client Group Profile”. This is where we will enter the NMS server IP(s).
Name – give it a name
Associated Management EPG – choose “default (Out-of-Band)”
Click the “+” sign under “Client Entries”.
Add the NMS servers (host) name and IP address.
Hit Submit again.
This completes this section.
Now Goto Fabric > Fabric Policies > Pod Policies > Policy Group >[right click] Create Pod Policy Group. You will see the below screen and fill it according to the sample shown below.
Name – give it a name
SNMP Policy – select the SNMP policy that you created earlier. In our example we created SNMP-Policy.
Now Goto Fabric > Fabric Policies > Pod Policies > Profile > Pod Profile default > default. You will see the below screen and fill it according to the sample shown below.
Choose the “Policy Group” that you created in the previous step. In our example it is “Fabric-Pod-Policy”.
This concludes the configuration that needs to be performed under Pod Policies.
Next we will configure under the Monitoring Policies.
2. Monitoring Policies
Goto Fabric > Fabric Policies > Monitoring Policies > Common Policy > Callhome/SNMP/Syslog
This is to allow objects in MIT to send traps to the NMS station.
In the work pane select SNMP and then from the tools icon ( ) select “Create SNMP Source”.
Name – give it a name (e.g. SNMP-Source-Objects)
Dest Group – Use the previously created destination group. In our example we used “SNMP-Traps-Receivers”
Configuring Monitoring Policies
Administrators can create monitoring policies with the following four broad scopes:
- Fabric Wide (Common): includes both fabric and access objects
- Access (infra): access ports, FEX, VM controllers, and so on
- Fabric: fabric ports, cards, chassis, fans, and so on
- Tenant: EPGs, application profiles, services, and so on
In our example we configured Fabric Wide (Common) Policies. You could also configure:
-Fabric policies (under Fabric > Fabric Policies > Monitoring Policies > default)
-Access policies (under Fabric > Access Policies > Monitoring Policies > default)
-Tenant policies (under Tenant > your-tenant-name > Monitoring Policies > [right click] Create Monitoring Policy
NOTE: To access context specific MIBs, you have to enable the tenant by right clicking on the VRF in the tenant and clicking “Create SNMP Context”. See below figure
3 thoughts on “Configuring SNMP for ACI”
Hi, for tenant SNMP monitoring policies, can i create it in common tenant, then apply it at the respective Tenants just like what you did with Syslog monitoring policies. In yes, then i don’t have to create the monitoring policy at each individual Tenant..
Is there a way to validate SNMP traps are being sent? i.e. is there a way to generate an ad hoc trap?