Veriflow this week introduced CloudPredict, a platform that provides organizations with clearer insights into their public clouds. The company supported the release with research suggesting that public cloud visibility indeed is lacking.
CloudPredict provides what the company calls “a consistent, established feedback loop” that equalizes insight into public clouds as well as private clouds and on-premise infrastructure. The software-as-a-service (SaaS) platform has four basic capabilities:
- Visualizing and searching paths and objects in public and multi-cloud environments
- Verifying and segmenting cloud applications enabling prediction of vulnerabilities
- Tracking changes and assuring compliance on-premise and in single and multi-cloud environments
- Ensuring what Veriflow calls “end-to-end reachability” across multi-cloud and hybrid environments
It Starts with Data
Organizational use of public clouds raises unique issues. Access to the data necessary to fully enfranchise operational and security elements can be difficult to get. Though the hardware in a public cloud is the responsibility of the provider, tenants are in charge of their own configurations, Brighten Godfrey, CTO and co-founder of Veriflow, told SDxCentral.
This divide can be tricky. “Unfortunately, the cloud has been like a black box to many networking teams: the provider doesn’t make it easy to see what happens in the virtual network layer, and solutions to visualize and verify cloud and multi-cloud networks have not kept pace with the increase in complexity as enterprises come to rely more on public cloud,” Godfrey wrote in an email.
A second issue is that multiple business groups – DevOps, networking, cloud and security – rely on the cloud. Keeping all these stakeholders informed and happy is not easy. Performance assessments and changes must be automated and quick. And this doesn’t always happen in public cloud environments.
Godfrey said that these environments encompass an increasing number of components (multiple clouds, virtual networks, interconnections, and others) that must fit together in seamless and secure ways.
Finally, public clouds are not all created equal. “There are also differences in the functional behavior of each cloud provider, many of which are not obvious,” Godfrey wrote. “A visibility and assurance solution needs to provide deeply accurate modeling of each cloud provider, but must present it in a unified provider-independent way to users.”
All told, it’s a dicey environment. Godfrey wrote that other platforms that assess public cloud performance deal with this by reactively monitoring applications and traffic.
CloudPredict, on the other hand, is intent-based. “It understands what could occur before it does, and determines whether that meets the intent,” he wrote. “For example, it can spot a vulnerability before it is exploited [by using] new mathematical analysis technology that just a few years ago was the subject of advanced academic research and is now available for networking teams.”
451 Research Senior Analyst Mike Fratto told SDxCentral that CloudPredict targets “a blind spot” by understanding and verifying cloud configurations and networking. Organizations do this via workflows, but Veriflow is adding value by providing “an automated and independent discovery, visualization, and verification to parts of the organization that may not be expert in cloud features.”
The question is whether existing ways of validating cloud performance will suffice to the extent that organizations don’t need the level of functionality Veriflow claims to offer. “For the enterprise, without a clear-cut mandate for external configuration validation such as regulatory requirements, then analytic software like this tends to fall into a non-critical bucket because the same goals can be accomplished via operational workflows,” Fratto said. “I think it’s less important to enterprises and not critical to use cloud services successfully.”
The research Veriflow released was conducted by Dimensional Research. It found 97 percent of responding companies reported problems with deploying and managing public clouds. The research revealed that IT teams disagree on “who is responsible, who approves network changes, whether the networking team should be involved in new deployments, and [the] perception of networking teams by cloud teams.”