3.7 Construct Xpath notation for a given node or instance of a node
Nodes
What is an XPath node? Basically, everything pertaining to a single XML document, starting from the root element. The primary XPath building block is a node. There are seven node types defined for XPath, [more info here](XPath Nodes.
Element
Attribute
Text
Namespace
Processing Instruction
Comment
Document
Every XML document is a tree of nodes, they all start from a root element. Between the nodes the connectivity is defined by various types of relationships.
Parent: Each node besides the root element has a parent node. Each node can have only one parent node.
Children: Each node, including the root element, may have one or more children.
Siblings: All the children of the same parent are the siblings to each other.
Ancestors: All the parent nodes up to the root element (for example, parent, parent’s parent) are ancestors.
Descendants: All the nested children down to the all leafs (for example, children, children’s children) are descendants.
When we speak of nodes, we must not think of network devices, as that was my original thought before I started to research this topic.
Now, there are a lot of concepts that we must know about XML, xpath to make sense. We must understand that all XML files start with a root element. This is a requirement. All documents start with this root element and form a Tree structure, by branching off into child elements which can then have more sub/child elements.
Another important term to get familiar with is Namespaces. XML provides Namespaces to avoid conflicts in duplicate nodes. Most likely namespaces carry a prefix behind it, like so:
<prefix:duplicate-nodename></prefix:duplicate-nodename>
<different-prefix:duplicate-nodename></different-prefix:duplicate-nodename>
Typically, namespaces will carry xmnls attributes about the namespace, with a URI declaration.
<prefix:duplicate-nodename xmlns:f="https://www.spauto.com/node-information"></prefix:duplicate-nodename>
There are a lot of different terms to learn about XML, a refresher is always good. Some information can be found at W3 - XML.
Why XPath
I’ve worked at places where some of the core routers have easily crossed over 100,000 lines of configuration. The configuration adds up quick, when we are talking about hundreds of L3/L2VPNs, MPLS, Traffic Engineering, Segment Routing, BGP, dozens of line-cards on Service based routers such as Nokia 7x50’s or Cisco 8000s. Using xpath is a simple way to filter through all the noise and only get what you need out of all the different elements in an incredibly large (or not) configuration.
NSO
If you want to use the NSO available in this repository, feel free to do so to try this out.
You will need to have loaded device and synced them with NSO, before you access the configuration of the routers. NSO provides a output modifier to return a command as an xpath expression.
The following example is from Devnet Labs
developer@ncs# show running-config devices device core-rtr01 config interface GigabitEthernet 0/0/0/1 ipv4 address | display xpath
/devices/device[name='core-rtr01']/config/cisco-ios-xr:interface/GigabitEthernet[id='0/0/0/1']/ipv4/address/ip 172.16.253.2
/devices/device[name='core-rtr01']/config/cisco-ios-xr:interface/GigabitEthernet[id='0/0/0/1']/ipv4/address/mask 255.255.255.248
developer@ncs#show
There is an additional tool to evaluate XPath expressions against the CDB data of NSO. You must enable it by turning on the devtools
developer@ncs# devtools true
Now you can use xpath eval.
XPath Expressions
We will be using XPath expressions in a way that an engineer would actually use them while crafting NETCONF rpc calls. A simple containerlab topology is provided to spin up an iosxr device.
All of this material will be stored in a folder called xpath_scripts in our project folder.
We’ll be using a NETCONF library that I developed for Nornir, Nornir Netconf
Lets get started!
TODO: