Ivan Pepelnjak, writing in IPSpace, heard from a reader who referenced Elisa Jasinska's presentation about network...
automation. In her presentation, Jasinska suggested that for a piece of network data, there should be "one source of truth."
The reader noted it's easy to think of multioperator scenarios where data sources such as an IP address management (IPAM) tool and Git source control get out of alignment. They added that when using Git to source control device configurations, data is often doubled up from IPAM. Network engineers have the option of storing IP and network device configuration data in source files, managed with Git, but this may complicate data access.
Pepelnjak weighed in on the subject of network device configuration data, agreeing with Jasinska that there should be a single source of truth for any form of data, and that data should be easily consumable, such as through an API. He added that IPAM data should be the single source of truth for network device configuration data, such as interface names, subnets and device names.
Pepelnjak recommended source-control tools like Git to manage device configuration templates and YAML files for auxiliary data. "While you're collecting device configurations and keeping track of changes in another Git repository, that's just a safeguard and a convenient method of identifying out-of-process changes made directly on the devices," he added.
Explore more of Pepelnjak's thoughts on network device configuration.
Looking to the next generation of network engineers
Tom Ammon, writing in Packet Pushers, said he believes in an age where the network is essential to business-critical applications, network engineers often treat other IT professionals -- such as server administrators or database administrators -- with disdain.
"The next generation of networking engineers will have to be much more skilled in their interactions with other IT professionals," Ammon said, pointing out that customers now expect tasks that once took weeks to be completed in minutes. This can only be achieved through an interdependent IT environment, where network engineers now borrow techniques once familiar only to server administrators.
According to Ammon, while network engineers often focus on "hard" skills for the future, such as building Linux boxes or troubleshooting data center fabrics, soft skills matter just as much. He said he believes empathy and persuasion will become critical.
"Empathy is what allows a talented engineer to sort through the emotions and irrelevant technical details and find the often small nugget of critical data that will lead to a real solution," Ammon said. "The next-generation network engineer will consider persuasion just as crucial a skill as interpreting a routing table," he added.
Dig deeper into Ammon's thoughts on network engineering.
Scaling up cybersecurity
Jon Oltsik, an analyst with Enterprise Strategy Group Inc., in Milford, Mass., said he believes cybersecurity is about to hit "an exponential curve" that will expand it well beyond the lofty promises of scale from security information and event management vendors.
Oltsik cited ESG research, which attributes the trend to more workloads moving into public and private clouds, the mainstreaming of containers and forecasts that call for more than 20 billion devices to be network-enabled by 2020. Other drivers may include expansions in physical networking from 10 Gb to 40 or 100 Gb, cellular network upgrades and the transition to IPv6 -- all representative of the greater digital transformation overall.
Oltsik said a scaling up of cybersecurity will drive other trends. These may include distributed data management, analytics and operations becoming entrenched as industry applications. He added that enterprise-class features like tiered administration, role-based access controls and centralized or distributed configuration options will also grow in importance.
Read more of Oltsik's thoughts on scaling up cybersecurity.
Getting a sense of network configuration tools
IT generalists and the future of network engineering
Scaling security with Google Container Engine