Home Cloud Computing Find out how to Use Ansible with CML

Find out how to Use Ansible with CML

Find out how to Use Ansible with CML


How can Ansible assist folks constructing simulations with Cisco Modeling Labs (CML)?

Much like Terraform, Ansible is a standard, open-source automation software usually utilized in Steady Integration/Steady Deployment (CI/CD) DevOps methodologies. They’re each a sort of Infrastructure as Code (IaC) or Infrastructure as Knowledge that let you render your infrastructure as textual content recordsdata and management it utilizing instruments resembling Git. The benefit is reproducibility, consistency, velocity, and the data that, if you change the code, folks approve, and it will get examined earlier than it’s pushed out to your manufacturing community. This paradigm permits enterprises to run their community infrastructure in the identical approach they run their software program and cloud practices. Afterall, the infrastructure is there to help the apps, so why handle them in another way? 

Though overlaps exist within the capabilities of Terraform and Ansible, they’re very complementary. Whereas Terraform is best on the preliminary deployment and making certain ongoing consistency of the underlying infrastructure, Ansible is best on the preliminary configuration and ongoing administration of the issues that reside in that infrastructure, resembling programs, community units, and so forth. 

In a standard workflow during which an operator desires to make a change to the community, let’s say including a brand new community to be marketed through BGP, a community engineer would specify that change within the code or extra seemingly as configuration information in YAML or JSON. In a typical CI workflow, that change would should be authorised by others for correctness or adherence to company and safety considerations, as an example. Along with the eyeball exams, a collection of automated testing validates the info after which deploys the proposed change in a take a look at community. These exams might be run in a bodily take a look at community, a digital take a look at community, or a mixture of the 2. That circulate would possibly appear to be the next: 

Workflow example

The benefit of leveraging digital take a look at networks is profound. The fee is dramatically decrease, and the power to automate testing is elevated considerably. For instance, a community engineer can spin up and configure a brand new, complicated topology a number of occasions with out the chance of previous exams messing up the accuracy of the present testing. Cisco Modeling Labs is a good software for such a take a look at. 

Right here’s the place the Ansible CML Assortment is available in. Much like the CML Terraform integration coated in a earlier weblog, the Ansible CML Assortment can automate the deployment of topologies in CML for testing. The Ansible CML Assortment has modules to create, begin, and cease a topology and the hosts inside it, however extra importantly, it has a dynamic stock plugin for getting details about the topology. That is necessary for automation as a result of topologies can change. Or a number of topologies might exist, relying on the exams being carried out. In case your topology makes use of dynamic host configuration protocol (DHCP) and/or CML’s PATty performance, the knowledge for a way Ansible communicates with the nodes must be communicated to the playbook. 

Let’s go over a few of the options of the Ansible CML Assortment’s dynamic stock plugin. 

First, we have to set up the gathering: 

ansible-galaxy assortment set up cisco.cml 

Subsequent, we create a cml.yml within the stock with the next contents to inform Ansible to make use of the Ansible CML Assortment’s dynamic stock plugin: 

plugin: cisco.cml.cml_inventory 

group_tags: community, ios, nxos, router

As well as to specifying the plugin identify, we will additionally outline tags that, when discovered on the units within the topology, add that machine to an Ansible group for use later within the playbook: 

As well as to specifying the plugin identify, we will additionally outline tags that, when discovered on the units within the topology, add that machine to an Ansible group for use later within the playbook:

  • CML_USERNAME: Username for the CML person
  • CML_PASSWORD: Password for the CML person
  • CML_HOST: The CML host
  • CML_LAB: The identify of the lab 

As soon as the plugin is aware of find out how to talk with the CML server and which lab to make use of, it could actually return details about the nodes within the lab: 

okay: [hq-rtr1] => { 

    "cml_facts": { 

        "config": "hostname hq-rtr1nvrf definition Mgmt-intfn!naddress-family ipv4nexit-address-familyn!naddress-family ipv6nexit-address-familyn!nusername admin privilege 15 secret 0 adminncdp runnno aaa new-modelnip domain-name mdd.cisco.comn!ninterface GigabitEthernet1nvrf forwarding Mgmt-intfnip handle dhcpnnegotiation autonno cdp enablenno shutdownn!ninterface GigabitEthernet2ncdp enablen!ninterface GigabitEthernet3ncdp enablen!ninterface GigabitEthernet4ncdp enablen!nip http servernip http secure-servernip http max-connections 2n!nip ssh time-out 60nip ssh model 2nip ssh server algorithm encryption aes128-ctr aes192-ctr aes256-ctrnip ssh shopper algorithm encryption aes128-ctr aes192-ctr aes256-ctrn!nline vty 0 4nexec-timeout 30 0nabsolute-timeout 60nsession-limit 16nlogin localntransport enter sshn!nend", 

        "cpus": 1, 

        "data_volume": null, 

        "image_definition": null, 

        "interfaces": [ 


                "ipv4_addresses": null, 

                "ipv6_addresses": null, 

                "mac_address": null, 

                "name": "Loopback0", 

                "state": "STARTED" 



                "ipv4_addresses": [ 



                "ipv6_addresses": [], 

                "mac_address": "52:54:00:13:51:66", 

                "identify": "GigabitEthernet1", 

                "state": "STARTED" 



        "node_definition": "csr1000v", 

        "ram": 3072, 

        "state": "BOOTED" 



The primary IPv4 handle discovered (so as of the interfaces) is used as `ansible_host` to allow the playbook to connect with the machine. We will use the cisco.cml.stock playbook included within the assortment to point out the stock. On this case, we solely specify that we would like units which might be within the “router” group created by the stock plugin as knowledgeable by the tags on the units: 

mdd % ansible-playbook cisco.cml.stock --limit=router 

okay: [hq-rtr1] => { 

    "msg": "Node: hq-rtr1(csr1000v), State: BOOTED, Handle:" 


okay: [hq-rtr2] => { 

    "msg": "Node: hq-rtr2(csr1000v), State: BOOTED, Handle:" 


okay: [site1-rtr1] => { 

    "msg": "Node: site1-rtr1(csr1000v), State: BOOTED, Handle:" 


okay: [site2-rtr1] => { 

    "msg": "Node: site2-rtr1(csr1000v), State: BOOTED, Handle:" 


Along with group tags, the CML dynamic stock plugin may even parse tags to cross data from PATty and to create generic stock details: 


If a CML tag is specified that matches `^pat:(?:tcp|udp)?:?(d+):(d+)`, the CML server handle (versus the primary IPv4 handle discovered) shall be used for `ansible_host`. To alter `ansible_port` to level to the translated SSH port, the tag `ansible:ansible_port=2020` might be set. These two tags inform the Ansible playbook to connect with port 2020 of the CML server to automate the desired host within the topology. The `ansible:` tag may also be used to specify different host details. For instance, the tag `ansible:nso_api_port=2021` can be utilized to inform the playbook the port to make use of to achieve the Cisco NSO API. Any arbitrary reality might be set on this approach. 

Getting began 

Attempting out the CML Ansible Assortment is straightforward. You should use the playbooks offered within the assortment to load and begin a topology in your CML server. To begin, outline the surroundings variable that tells the gathering find out how to entry your CML server: 

% export CML_HOST=my-cml-server.my-domain.com 

% export CML_USERNAME=my-cml-username 

% export CML_PASSWORD=my-cml-password 

The following step is to outline your topology file. That is a normal topology file you can export from CML. There are two methods to outline the topology file. First, you’ll be able to use  an surroundings variable: 

% export CML_LAB=my-cml-labfile 

Alternatively, you’ll be able to specify the topology file if you run the playbook as an additionalvar.  For instance, to spin up a topology utilizing the inbuilt cisco.cml.construct playbook: 

% ansible-playbook cisco.cml.construct -e wait="sure" -e  

This command masses and begins the topology; then it waits till all nodes are operating to finish.  If -e startup=’host’ is specified, the playbook will begin every host individually versus beginning them all of sudden.  This enables for the config to be generated and fed into the host on startup.  When cml_config_file is outlined within the host’s stock, it’s parsed as a Jinja file and fed into that host as config at startup.  This enables for just-in-time configuration to happen. 

As soon as the playbook completes, you should use one other built-in playbook, cisco.cml.stock, to get the stock for the topology.  With a purpose to use it, first create a cml.yml within the stock listing as proven above, then run the playbook as follows: 

% ansible-playbook cisco.cml.stock 

PLAY [cml_hosts] ********************************************************************** 

TASK [debug] ********************************************************************** 

okay: [WAN-rtr1] => { 

    "msg": "Node: WAN-rtr1(csr1000v), State: BOOTED, Handle:" 


okay: [nso1] => { 

    "msg": "Node: nso1(ubuntu), State: BOOTED, Handle: my-cml-server.my-domain.com:2010" 


okay: [site1-host1] => { 

    "msg": "Node: site1-host1(ubuntu), State: BOOTED, Handle: site1-host1:22" 


On this truncated output, three completely different eventualities are proven.  First, WAN-rtr1 is assigned the DHCP handle it obtained for its ansible_host worth, and ansible port is 22. If the host operating the playbook has IP connectivity (both within the topology or a community linked to the topology with an exterior connector), will probably be in a position to attain that host. 

The second state of affairs reveals an instance of the PATty performance with the host nso1 during which the dynamic stock plugin reads these tags to find out that the host is out there via the CML server’s interface (i.e. ansible_host is ready to my-cml-server.my-domain.com).  Additionally, it is aware of that ansible_port needs to be set to the port specified within the tags (i.e. 2010).  After these values are set, the ansible playbook can attain the host within the topology utilizing the PATty performance in CML. 

The final instance, site1-host1, reveals the state of affairs during which the CML dynamic stock script can both discover a DHCP allotted handle or tags to specify to what ansible_host needs to be set, so it makes use of the node identify.  For the playbook to achieve these hosts, it must have IP connectivity and be capable to resolve the node identify to an IP handle. 

These built-in playbooks present examples of find out how to use the performance within the CML Ansible Assortment to construct your individual playbooks, however you may as well use them straight as a part of your pipeline.  In truth, we frequently use them straight within the pipelines we construct for patrons. 

If you wish to be taught extra in regards to the CML Ansible Assortment, yow will discover it in Ansible Galaxy in addition to on Github. 

It’s also possible to discover a full, IaC CI/CD pipeline utilizing these modules right here.


Be a part of the Cisco Studying Community at the moment without cost.

Comply with Cisco Studying & Certifications

Twitter | Fb | LinkedIn | Instagram | YouTube

Use #CiscoCert to hitch the dialog.




Please enter your comment!
Please enter your name here