In this Help Net Security interview, Lingping Gao, CEO at NetBrain, discusses the challenges NetOps teams face in maintaining production services due to outdated processes and growing infrastructures.
No-code automation has the potential to address these challenges by allowing engineers to automate repetitive tasks, thereby enhancing efficiency and mitigating the risk of failures.
What are the current challenges NetOps teams face in managing network infrastructures?
Simply put, NetOps teams are pushed to their limits and falling behind to maintain normal production services. These teams are consumed by remedial tasks and the repetitive operations required to address and close user service requests. It’s a problem of scale. As infrastructure grows and more mission-critical services depend on that infrastructure, service requests skyrocket.
Using the manual approach to restorative services requires staff to grow as request volume grows – but t is no longer supported by the macroeconomics of the enterprise. Repair times increase, and the number of reported incidents climbs. And even worse, as this problem grows, so does the risk of more catastrophic failures which will ultimately affect customer confidence.
Outdated processes hold NetOps teams back. In fact, the processes most NetOps teams used 25 years ago are still in use today: relying on manual command-line sequences and individual ad-hoc scripts. This bespoke approach becomes the standard for all engineers who are unable to see that there is a better way — one that automates a large portion of the repetitive work and allows the subject matter experts to do what they do best — solve hard problems.
Considering the increasing importance of cybersecurity, what are the key benefits and challenges of merging network operations (NetOps) with security operations (SecOps)?
At its core, infrastructure security is a function of the network. Everything that contains information is attached to the network, and these devices are accessed internally and externally. This means that the attack surfaces and hardening of the infrastructure is fundamentally defined by the SecOps teams and then largely implemented by the NetOps team. For instance, the SecOps team may define the requirement to coordinate firewall policies, but the NetOps team is best suited to do so. The SecOps teams may define access rules, port availability, password compliance and firmware versions that are required, but it is the NetOps teams that are best suited to provide this design and operational status assessment automatically and on a regular basis.
With the rise of AIOps, how does it enhance network performance management, particularly in dealing with data overload and complex infrastructures, and what potential solutions does it offer?
AIOps has been in development for decades and has the promise of being able to identify solutions. The issue is most NetOps teams already know the solutions that are needed, they just can’t scale the people with that knowledge to the entire organization. So, while AIOps may find a home in the future, the knowledge available within subject matter experts already far exceeds anything AIOps will bring.
When NetOps couples the ability to capture SME knowledge (through no-code) with the ability to execute that knowledge (by emulating the SME themselves), the need for AIOps for networks is diminished. Network automation can emulate SMEs at scale, making it a better fit for this problem.
What are the common starting points for organizations looking to implement network automation, and how can they scale these efforts over time?
Organizations should start by looking for any tasks that are repeated regularly. It’s okay to think small at first. The biggest mistake organizations can make in the beginning is believing that they must boil the ocean and spend far too long (typically measured in years) trying to formulate a ‘master plan.’ This prevents automation from being deployed in the short term and creates rigid structures that fail to deliver the needed value by the time those long lead-time projects are completed. By starting with simple, basic and scalable security needs, automation platforms can deliver value and become more adopted internally.
A great example is using automation to continuously assess the firmware versions of the organization’s firewalls. Manufacturers update this firmware regularly, and it’s quite easy for an overworked engineer to ‘forget’ to update one or more of the many firewalls in production. Some of the biggest breaches in recent years have been attributed to errant/dated firmware versions.
What future trends do you foresee in network automation and NetOps?
One trend I’ve seen is the use of no-code/low-code, which essentially democratizes automation — making it available and accessible to every network engineer. Simply put, if an engineer can think of something that they do repetitively, then they can capture that task into automation using no-code and thereafter let the machine do that work automatically.
No-code automation eliminates bottlenecks and allows non-technical staff to automate ANY NetOps procedures that could otherwise only be done by a few network engineers who know Python or Ansible. It has many uses in NetOps, from outage prevention to network security, including automating common diagnostic and troubleshooting tests, identifying configuration drift, network discovery and mapping, and testing if security products are operating as expected. It empowers non-technical business users to build enterprise-grade applications and processes. With no-code, engineers can think about using automation FIRST to eliminate a large portion of the tedium, and then focus their attention on the harder part of their job – fixing the problem.