Impact of Public Cloud Adoption: DevOps and Network Teams at Odds

Moving the cloud always sounds like a good idea to the DevOps team, but what about the network engineers?

In 2019, enterprises are expected to invest $214 billion in cloud services and apps, up 17.5 percent from 2018. Most initial public cloud deployments started as pet projects that DevOps teams adopted to make their enterprise processes more Agile. However, cloud and DevOps teams are increasingly scaling deployments for more mission-critical business-wide use cases. The underlying network infrastructure is becoming increasingly complex with the need to provide network-wide assurance for multi-cloud environments.

Many networking teams are having a hard time keeping up with demands. This is putting networking/operations and cloud/DevOps teams at odds.

Cloud teams are rapidly porting new applications to the cloud without running changes through network teams. However, network teams are still ultimately responsible for ensuring network uptime as well as availability — both on-premises and in cloud environments. And while there are solutions to monitor cloud applications or virtualization layers, the network layer too often goes disregarded, leaving questions of visibility and assurance that ultimately become an obstacle to overall business goals.

To understand more about the challenges associated with public cloud adoption, Dimensional Research recently released a report that surveyed more than 300 respondents from networking, operations, cloud, and DevOps disciplines. The findings revealed three key problems with public cloud adoption: Public cloud adoption creates unforeseen turmoil across networking, operations, cloud and DevOps teams; This unforeseen turmoil impacts productivity in multiple areas; and, the impacted productivity ultimately leads to negative financial impacts to companies.

Public Cloud Turmoil: The Battleground of Networking/Operations vs. Cloud/DevOps

The divide within today’s IT teams starts with a disconnect between who is ultimately responsible for the network in public cloud deployments. On one hand, networking and operations teams believe they are responsible for the public cloud network, as the network has historically been their domain. On the other hand, cloud and DevOps teams are expanding their deployments and are of the opinion that cloud, networking, and security teams should have a shared stake in network ownership and oversight. This poses a great risk to companies in a few areas.

In change management, these teams cannot agree on who has the final say. The vast majority (90 percent) of networking teams view it as their responsibility and theirs alone, while two-thirds of cloud teams think this is their responsibility.

New public cloud deployments tell a similar story. Cloud and DevOps teams often opt to not include networking teams when they deploy new cloud environments. Dimensional Research’s findings reveal that almost 40 percent of cloud and DevOps teams push forth with new cloud deployments without involving networking teams. This is a slippery slope and can result in a slew of compliance issues and security challenges.

With that said, cloud teams are not the bad guys here. Their primary goal is to make the businesses they serve more Agile and many believe that networking teams cannot keep up. Indeed, Dimensional Research found that 65 percent of cloud and DevOps teams view their networking teams as “too slow” for more Agile cloud environments. And with the rapid pace of operations in today’s enterprises, being “too slow” can be negatively impact a business’s bottom line.

What’s the magic fix here? The missing key here is visibility. When networking and operations teams are unaware of changes made to the network and/or new deployments implemented, they do not have the means to perform due diligence checks for proper segmentation and compliance. This puts the company at risk for network downtime and vulnerabilities. And ultimately, a downed network is the costliest problem of all.

Productivity: The Collateral Damage from Public Cloud Turmoil

Despite their differences, networking, operations, cloud, and DevOps teams all agree that the public cloud is impacting productivity in multiple areas, including: deployment, management, and security and compliance.

With half of all public cloud environments deployed without any involvement from networking teams, IT teams are operating with one hand tied behind their backs from the get-go. This becomes especially apparent in the management of public cloud environments. Despite the vast majority of IT teams (92 percent) prioritizing public cloud infrastructure as important or very important, approximately 80 percent have less visibility into the public cloud portion of their network. Furthermore, many of them do not have solutions that can provide easy and direct visibility into the virtual networks of their public cloud environments.

Lack of deployment and management leads to more serious issues such as overall network security and compliance. Despite the public cloud’s potential to make enterprises more Agile, improper use leads to increased security threats. More than half of all IT teams report an increase in the frequency of security threats due to public cloud usage, and more than 60 percent are unable to verify segmentation of their network. This trickles downs to compliance issues as well – a very small proportion of public cloud environments are compliant with industry standards and regulations.

To reestablish productivity, networking and cloud teams need to get on the same page with deploying and managing public cloud networks. Much like with network visibility, proactive assurance and security makes public cloud deployments much more manageable for all involved parties. With that said, since assurance and security have historically fallen under the networking team’s purview, typically only CCIEs are fluent in the languages and processes needed for network assurance. The way to remedy this is by deploying intent-based networking solutions that the most technical network engineers and the most innovative cloud developers can use.

Public Cloud Problems Are Financial Problems

As noted, public cloud adoption is growing at an exponential rate – unfortunately, the frequency of problems is growing just as rapidly. Around half of IT teams have at least monthly problems within their public cloud implementations, with 20 percent coming across problems on a weekly basis. Time to resolution exacerbates this issue, as more than half of these teams say it takes three hours or more to resolve issues in their public cloud networks.

Once the network goes down, any one (or all) of four things can happen. First, the cost of operations skyrockets, as almost 50 percent of IT teams report increased resource costs associated with remediation of public cloud downtime. Second, revenue takes a plunge, as almost half of IT teams also cite revenue loss as a direct result of failures in their public cloud networks. Third, customer satisfaction follows suit, as nearly 70 percent of IT teams see a decrease in customer satisfaction. And fourth, the brand of the company is tarnished (40 percent).

Where Do We Go from Here?

At the end of the day, the one thing networking, operations, cloud, and DevOps teams agree on is the need for a better solution. Key requirements across all teams include clear visibility across public cloud deployments and the ability to predict issues in the public cloud before they happen. This starts with getting networking teams and cloud teams on the same page when any and all new public cloud environments are deployed, and when any changes are made. From there, these two disparate teams need to find common ground in the management of public cloud networks, with solutions that work for both ends. There are many moving parts within each public cloud deployment – having a solution that can proactively predict vulnerabilities and provide visibility for both parties is becoming increasingly important for public cloud management.

Then, and only then, can IT teams collectively achieve sustainable agility in their public cloud deployments, which will ultimately drive toward their intended business goal.

Impact of Public Cloud Adoption: DevOps and Network Teams at Odds
DZone – April 19, 2019