Keeping Virtual Security Threats From Blindsiding Your Network
By Mark Boltz, CISA, CISSP, NSA-IEM, CSGI
Senior Solutions Architect
Stonesoft Inc.
A few weeks ago, I was speaking with a customer whose business was expanding fast, so he had to add three new servers to his server farm. To save precious floor space, he decided to virtualize a couple of these servers. The results were so encouraging that he virtualized the rest of the servers that had a low utilization rate. Again, everything went fine, and he saved enough floor space so that there would still be room for a dozen additional servers.
What was odd, though, was that the customer did not receive as many firewall or Intrusion Prevention System (IPS) alerts as he used to. It was as if a large chunk of network traffic had just disappeared. After checking all of the firewall, IPS and network monitoring logs, and finding them in normal condition, he realized that his new virtual environment was a blind spot for traditional network security appliances -- they could not see what happened there.
This customer is representative of the growing trend in the data center today. Virtualization makes it possible to deploy multiple virtual servers, each running separate operating systems and applications on one physical server. The results are more efficient usage of existing hardware, reduced power and cooling costs and reduction in data center footprints…and a complex one-stop shop for hackers trying to access valuable data. This is definitely not an ideal location for a "blind spot" on the network. The customer quickly realized that this wouldn't be nearly as plain and simple as he expected.
The task at hand
Virtualization presents a number of challenges to keeping the data center secure -- the first of which is complexity.
The nature of virtual environments -- multiple software architectures running in a single physical server -- gives many organizations a false sense of security. Although the physical architecture is different, virtual machines are still running typical operating systems and applications that require constant patches and updates to remain secure. A virtual environment is just as vulnerable to attacks as any other device connected to the network.
In addition, as with the customer's situation above, when one virtual appliance communicated with another virtual appliance, the communication never left the virtual environment. In other words, it never went through any network security devices that were outside of the virtual environment. So, the network security devices could not see what was happening inside the virtual environment. Not only does this present a major hole in security architecture, it is also not compliant with governance regulations.
To ensure the integrity of data, all virtual networks must be secured from one another. This means preventing threats from entering the physical server and blocking malicious traffic from moving between the virtual machines that it hosts.
A variety of options to consider
While there are many different approaches to securing virtual environments, such as retrofitting existing hardware or virtual LAN (VLAN) tagging, many of today's options have less than optimal ramifications.
VLAN tagging, or creating separate zones of trust, is very complex as each VLAN connection must have its own set of policies for acceptable use. In addition, these policies must be managed and updated manually, which creates a large potential for security risks.
Using hardware, such as traditional firewalls and IPS applications, in between each virtual host adds complexity to the network and creates even greater security and availability risks. For example, hardware-based solutions reside on an appliance and are located in front of the system they are securing. However, as the core of virtualization is the removal or reduction of servers and hardware, virtual environments will require this hardware to be eliminated. In this case, i.e., without the hardware, the security device is only able to throw a blanket of protection over the entire virtual environment, not over each individual component.
Perhaps the most important consideration here is the level of visibility. Traditional management consoles have not been designed to provide visibility into virtual environments. They cannot report the amount of traffic passing between virtual systems. Nor can they alert administrators if a system is about to go down or help to quickly address security threats.
In comparison, software-based security appliances are administered on a device where all the security functions are implemented in software that is run on top of or as a part of an operating system. The software-based virtual security appliance is very similar. However, in this case, the physical device is replaced with a virtual machine. In other words, software-based virtual security solutions are architected similarly to traditional appliances -- except that they do not need the hardware to operate.
With software-based virtual security solutions, organizations can gain a high level of visibility into virtual networks. In addition, there are some virtual solutions with management consoles that uniquely provide unified visibility across both virtual and physical networks. Since many organizations will evolve to virtual environments over time, this will be extremely important. As a result, IT departments have the real-time monitoring, management and analysis tools they need to efficiently ensure the continuous uptime and security of their networks. In addition, they have the auditing and reporting control needed for regulatory compliance.
Implementing virtual security solutions, such as virtual firewall/VPN, IPS and SSL VPN appliances, within the virtual environment enables organizations to maximize the benefits of virtualization with the confidence that their systems will be secure and always available. In addition, since virtual security "appliances" are inherently fully functional pieces of software, they have the flexibility to be deployed within any virtual environment. This unprecedented flexibility allows the organization's IT staff to focus on larger business issues, such as better risk mitigation, regulatory compliance, operational continuity and cost reduction.
In summary, software-based virtual security appliances are a great solution to the virtualization security dilemma. They are specifically designed for easy deployment in virtual environments to ensure maximum visibility, reduced complexity and minimal costs. However, most importantly, software-based virtual security solutions allow organizations to reap the benefits of virtualization without sacrificing visibility, security or compliance with industry regulations.
Failure to plan = planning to fail
It is clear that virtualization offers organizations significant benefits in the form of improved efficiency, lower energy costs and consolidation of data centers. On the other hand, without sufficient understanding of the network security tools available to protect against internal and external virtual threats, security professionals may find themselves blindsided by virtual security threats.
There is no benefit if network administrators adopt virtualization in effort to reap higher network agility and resource management, yet at the same time sacrifice network visibility and control that are required for effective security infrastructure management. In this case, failing to plan virtual security is nothing but planning to endure virtual security breaches. However, by implementing an appropriate virtual security strategy with professional tools, virtual networks can function with superior efficiency and without posing hindrances to the value of data center virtualization.