Submitted by: Adam Boyle, Head of Product Management, Hybrid Cloud Security, Trend Micro
When it comes to software container security, it’s important for enterprises to look at the big picture, taking into account how they see containers affecting their larger security requirements and future DevOps needs. Good practices can help security teams build a strategy that allows them to mitigate pipeline and runtime data breaches and threats without impacting the agility and speed of application DevOps teams.
Security and IT professionals need to address security gaps across agile and fast pace DevOps teams but are challenged by decentralized organizational structures and processes. And since workloads and environments are constantly changing, there’s no silver bullet when it comes to cybersecurity, there’s only the info we have right now. To help address the current security landscape, and where containers fit in, we need to ask ourselves a few key insightful questions.
How have environments for workloads changed and what are development teams focused on today? (i.e. VMs to cloud to serverless > DevOps, microservices, measured on delivery and uptime).
Many years ago, the customer conversations that we were having were primarily around cloud migration of traditional, legacy workloads from the data center to the cloud. While performing this “forklift,” they had to figure out what IT tools, including security, would operate naturally in the cloud. Many traditional tools they had already purchased previously, before the cloud migration, didn’t quite work out when expanded to the cloud, as they weren’t designed with the cloud in mind.
In the last few years, those same customers who migrated workloads to the cloud, started new projects and applications using cloud native services, and building these new capabilities on Docker, and serverless technologies such as AWS Lambda, Azure functions, and Google Cloud functions. These technologies have enabled teams to adopt DevOps practices where they can essentially continuously deliver “parts” of applications independently of one and other, ultimately delivering outcome much faster to market than one would with a monolithic application. The new projects have given birth to CI/CD pipelines leveraging Git for source code management (using hosted versions from either GitHub or BitBucket), Jenkins, or Bamboo for DevOps automation, and Kubernetes for automated deployment, scaling, and management of containers.
Both of these thrusts are now happening in parallel driving two distinct classes of applications—legacy, monolithic applications, and cloud native microservices. The questions for an enterprise are simple; how do I protect all of this? And, how can I do this at scale?
What’s worth mentioning is also the maturity of IT and how these teams have evolved into leveraging “infrastructure as code.” That is, writing code to automate IT operations. This includes security as code or writing code to automate security. Cloud operations teams have embraced automation and have partnered with application teams to help scale the automation of DevOps driven applications while meeting IT requirements. Technologies like Chef, Puppet, Ansible, Terraform, and Saltstack are popular in our customer base when automating IT operations.
While vulnerabilities and threats will always persist, what is the bigger impact on the organization when it comes to DevOps teams and security?
What we hear when companies talk to us is that the enterprise is not designed to do security at scale for a large set of DevOps teams who are continuously doing build->ship->run and need continuous and uninterrupted protection.
A typical enterprise has a centralized IT and Security Ops teams who are serving many groups of internal customers, typically business units which are responsible for generating the revenue for the enterprise.
So, how do tens or hundreds of DevOps teams who continuously build->ship->run, interact with centralized IT and security Ops teams, at scale? How do IT and security Ops teams embrace these practices and technologies, and ensure that they are secure—both the CI/CD pipelines and the runtime environments?
These relationships between IT teams (including security teams), and the business units have largely been at an executive level (VP and up), but to deliver “secure” outcomes continuously—a more effective, a more automated interplay—between these teams are needed.
We see many DevOps teams across business units incorporating security with varying degrees of rigor—or buying their own security solutions that only work for their set of projects—purchased out of their business unit budgets, implementing them with limited security experience and no tie-back to corporate security requirements or IT awareness. This leads to a fragmented, duplicated, complicated, inconsistent security posture across the enterprise and higher cost models on security tools that becomes more complicated to manage and support. The pressure to deliver faster within a business unit is sometimes at the cost of a coordinated enterprise-wide security plan…we’ve all been there and there’s often a balance that needs to be found.
The relationship, at the working level, between business unit application teams and centralized IT and security Ops teams is not always a collaborative, healthy, working relationship. Sometimes it has friction. Sometimes, the root cause of this friction can be related to application teams having significantly higher understanding of DevOps practices, tools, along with higher understanding of technologies, such as Docker, Kubernetes, and various serverless technologies, than their IT counterparts. We’ve seen painful, unproductive discussions between application teams trying to educate their IT/Security teams on the basics, let alone, get them on board with doing things differently. The friction increases if the IT and security Ops teams don’t embrace the changes in their approach when it comes to container and serverless security. So, to us, the biggest impact right now is if a DevOps team wants to deliver continuously while following an enterprise-wide approach, then they need a continuous relationship with the IT and security operations teams, whom must become well educated in DevOps practices and tools, and microservices technologies (Docker, Kubernetes, etc), where the teams work together to automate security across pipelines and runtime environments. And, the IT and security teams need to level up their skills sets to DevOps and all associated technologies, and help teams move faster, not slower, while meeting security requirements.
To be true DevOps, the “Dev” part would be the application team, the “Ops” part would be ideally IT/security and they would work together. So, we think there could be some pretty big shifts on how enterprises organize their development teams and IT/security Ops teams as the traditional organizational models favor delivery of monolithic, legacy applications that do not do continuous delivery.
The biggest opportunity for IT/security Ops teams is engage the application teams with a set of self-service tools and practices that are positioned to help the teams move faster, while meeting the IT and security requirements for the enterprise.
How can DevOps teams take advantage of the best security measures to better protect emerging technologies like container environments and their supporting tools?
Well this could easily be a book! However, let’s try to summarize at a high level and break this down into “build,” “ship,” and “run.” By no means is this a complete list, but enough to get started. For more information, contact us
Security teams have fantastic opportunity to introduce the following services across the enterprise, for all teams with pipelines and runtimes, in a consistent way.
- Identification of all source code repositories and CI/CD pipelines across the enterprise, and their owners.
- Static code analysis.
- Image scanning for malware.
- Image scanning for vulnerabilities.
- Image scanning for configuration assessments (ensure images are hardened).
- Indicator of Compromise (IoC) queries across all registries.
- Secrets detection.
- Automated security testing in staged environments, with generic and custom test suites.
- Image Assertion – declaring an image to be suitable for the next stage of the lifecycle based on the results of scans, tests, etc.
- Provide reporting to both application teams and security teams on security scorecards.
- Admission control – the allowance or blocking of images to runtime environments based on security policies, image assertion, and/or signed images.
- Vulnerability shielding of containers – Trend Micro will be releasing this capability later this year.
- Runtime protection of Docker and Kubernetes, including anomaly detection of abnormal changes or configurations.
- Hardening of Kubernetes and Docker.
- Using Kubernetes network policy capabilities for micro-segmentation, and not a third-party solution. Then, ensure Kubernetes is itself protected.
- Container host-based protection—covering malware, vulnerabilities, application control, integrity monitoring, and log inspection—for full stack defense of the applications and the host itself.
- Kubernetes pod-based protection (privileged container – one per pod). This can be shipped into Kubernetes environments just like any other container, and no host-based agent is required.
For serverless containers and serverless, application protection in every image or serverless function (AppSec library focusing on RASP, OWASP, malware, and vulnerabilities inside the application execution path). Trend Micro will be releasing an offer later this year to address this.
Trend Micro provides a stronger and more robust full lifecycle approach to container security. This approach helps application teams meet compliance and IT security requirements for continuous delivery in CI/CD pipelines and runtime environments. With multiple security capabilities, complete automation resources, and world class threat intelligence research teams, Trend Micro is a leader in the cybersecurity needs of today’s application and container driven organizations.
Learn more at www.trendmicro.com/containers.
The post The Next Enterprise Challenge: How Best to Secure Containers and Monolithic Apps Together, Company-wide appeared first on .