fabrics. It is possible to deploy SD-Access by using
automation tools and connecting to each device using
NETCONF, the CLI, or even SNMP. This gives the
network administrator more granular control but at the
same time requires more work and time to define
configuration parameters and data models and to ensure
that the correct configuration gets applied to all the
devices in a unified manner. Cisco DNA Center performs
all these tasks automatically, and automation scripts can
be used to collect operational data and ensure that the
network performs within optimal parameters.
INFRASTRUCTURE AS CODE
Inspired by software development practices,
infrastructure as code is a new approach to
infrastructure automation that focuses on consistent,
repeatable steps for provisioning, configuring, and
managing infrastructure. For a long time, infrastructure
has been provisioned and configured in a manual
fashion. Deploying a new application on a network used
to take anywhere from weeks to months to even years in
some situations. During that time, the IT team would try
to size the hardware requirements for the new
application, go through the approval process to purchase
the new hardware, order the hardware, wait for it to be
delivered, rack it up, and start the provisioning and
configuration steps. The provisioning and configuration
of the new hardware was in most cases a manual process:
installing the operating system one step at a time
following guided instructions and then finally installing
and configuring the new application. At best,
technologies like PXE booting and installation scripts
would be used to try to automate the arduous process of
provisioning new hardware and installing new
applications. This resulted in one-of-a-kind, or
“snowflake,” environments in which each application ran
on different hardware, with different options and
features enabled; supporting and maintaining such