In the last year, most organizations have accelerated adoption of cloud services far more rapidly than they originally planned to, but that doesn't mean that everything is moving to the cloud. Most organizations are adopting a hybrid approach to their cloud deployment; while traditional enterprises have operated on-premises for decades, they can clearly see the advantages of public cloud.
Cloud computing provides significant benefits to organizations, enabling far more agility for development teams, on-demand scalability to meet variable compute needs, and reduced responsibility for infrastructure and the staff required to maintain it.
But while cloud service providers (CSPs) — such as Amazon Web Services, Microsoft Azure, and Google Cloud Platform — are responsible for securing the infrastructure itself, the organizations themselves remain responsible for the security of their data and applications. Thus, the slower shift for established enterprises; they are moving workloads up to the cloud as they can, while still managing challenges related to rearchitecting applications, addressing compliance and security concerns, and finding the right talent and resources to manage their new hybrid environments.
Why Do Organizations Go Hybrid?
What organization today can operate successfully in just one environment? Not many. Neither a purely on-premises environment nor a single public cloud can truly address modern business needs because some operations work best in public cloud environments while others are better suited to private cloud and on-premises environments.
Hybrid cloud incorporates some level of workload portability, orchestration, and management across two or more environments, such as one (or more) public clouds, and one (or more) private clouds (which may be hosted on-premises or off), and bare-metal or virtual environments. In addition, many organizations are concerned about locking themselves into a single CSP and losing flexibility.
Each organization must achieve their own delicate balance among scale, speed, privacy, and security through a mix of services under a hybrid model.
Data Security Considerations for Hybrid Cloud Environments
The advantages of a hybrid deployment model are clear, but this model also creates a host of data management and data security challenges. Attackers don't care where a workload is hosted, and in fact, an attack might move from the on-premises local network up to the cloud, and vice versa. Security teams need visibility that spans all environments so they can understand the broader picture of what malicious actors might be doing.
To get true visibility into their environments, organizations need to be able to observe anything that could be considered an endpoint, which today could include PC workstations, Mac laptops, Amazon Elastic Compute Cloud (Amazon EC2) instances, virtual machines in clouds, and container workloads. There are so many types of computing endpoints spread across so many environments that most organizations struggle to gather security telemetry and build a unified view of all their endpoints.
The Complexity of Hybrid Cloud Security
As these organizations go through the transition to hybrid environments, they have a lot to consider as they plan their best path forward. Some newer enterprises are cloud native and assume that hybrid doesn't apply to them, yet they, too, have many endpoints to manage — the laptops of their remote workers are actually on-premises endpoints, and they have more cloud configurations and instances than they planned for or expected.
And while many traditional enterprises already have on-premises security tooling, it may not expand into the cloud as they adopt a hybrid model. Development teams find that what they use on premises doesn't cover their needs in a cloud environment, which brings developers and security teams back to a familiar problem: multiple tools, multiple outputs, and a lot of endpoints.
Having multiple tools introduces cost and management concerns, as well as more tools for staff to learn and manage. When it comes to detection and investigation, security professionals need solutions that they understand, enable them to consolidate results, and don't require them to switch tools when moving from validation to detection. The many different sources of data — from production endpoints and servers on-premises and cloud resources, workloads, containers, and SaaS data (generated by using G Suite, Slack, Jira, and many other tools) — increase the challenges of analyzing all of the different endpoint data.
In case of an attack, incident investigators need to get answers for many questions. Who else did this server talk to? Where else has this user logged in? What external connections has this server made? Where else has this file hash shown up? Telemetry can help investigators understand who is accessing the data, modifying it, or even exfiltrating it. Telemetry data can also help with insider threats, by making it easier to tell if an insider starts to access folders and files they usually don't touch.
What You Need to Manage Hybrid Cloud Security
Hybrid cloud security is complex because it spans so many environments. To manage and maintain security in hybrid environments, ensure that security teams can:
Together, all of this data empowers organizations to detect and investigate potential problems. Fragmented information and processes help attackers, because they lead to inevitable roadblocks in investigation as security teams wait for logs and telemetry services from other sources. Collecting and analyzing endpoint telemetry data provides organizations with a continuous and proactive security posture, while still making the most of the many important benefits hybrid cloud offers in an increasingly competitive marketplace.Ganesh Pai is Founder & CEO of Uptycs. He was previously Chief Architect, Carrier Products and Strategy for Akamai Technologies, a leading provider of content delivery network services. Prior to Akamai, Ganesh was founder & VP Systems Architecture of Verivue, a leading ... View Full Bio