Biblio

Edge computing promises higher bandwidth and lower latency to end-users. However, edge servers usually have limited computing resources and are geographically distributed over the edge. This imposes new challenges for efficient system monitoring and control of edge servers.In this paper, we propose EdgeVMI, a framework to monitor and control services running on edge servers with lightweight virtual machine introspection(VMI). The key of our technique is to run the monitor in a lightweight virtual machine which can leverage hardware events for monitoring memory read and writes. In addition, the small binary size and memory footprints of the monitor could reduce the start/stop time of service, the runtime overhead, as well as the deployment efforts.Inspired by unikernels, we build our monitor with only the necessary system modules, libraries, and functionalities of a specific monitor task. To reduce the security risk of the monitoring behavior, we separate the monitor into two isolated modules: one acts as a sensor to collect security information and another acts as an actuator to conduct control commands. Our evaluation shows the effectiveness and the efficiency of the monitoring system, with an average performance overhead of 2.7%.

In monolithic operating system (OS), any error of system software can be exploit to destroy the whole system. The situation becomes much more severe in cloud environment, when the kernel and the hypervisor share the same address space. The security of guest Virtual Machines (VMs), both sensitive data and vital code, can no longer be guaranteed, once the hypervisor is compromised. Therefore, it is essential to deploy some security approaches to secure VMs, regardless of the hypervisor is safe or not. Some approaches propose microhypervisor reducing attack surface, or a new software requiring a higher privilege level than hypervisor. In this paper, we propose a novel approach, named HyperPS, which separates the fundamental and crucial privilege into a new trusted environment in order to monitor hypervisor. A pivotal condition for HyperPS is that hypervisor must not be allowed to manipulate any security-sensitive system resources, such as page tables, system control registers, interaction between VM and hypervisor as well as VM memory mapping. Besides, HyperPS proposes a trusted environment which does not rely on any higher privilege than the hypervisor. We have implemented a prototype for KVM hypervisor on x86 platform with multiple VMs running Linux. KVM with HyperPS can be applied to current commercial cloud computing industry with portability. The security analysis shows that this approach can provide effective monitoring against attacks, and the performance evaluation confirms the efficiency of HyperPS.

The principle of least privilege requires that components of a program have access to only those resources necessary for their proper function. Defining proper function is a difficult task. Existing methods of privilege separation, like Control Flow Integrity and Software Fault Isolation, attempt to infer proper function by bridging the gaps between language abstractions and hardware capabilities. However, it is programmer intent that defines proper function, as the programmer writes the code that becomes law. Codifying programmer intent into policy is a promising way to capture proper function; however, often onerous policy creation can unnecessarily delay development and adoption. In this paper, we demonstrate the use of our ELF-based access control (ELFbac), a novel technique for policy definition and enforcement. ELFbac leverages the common programmer's existing mental model of scope, and allows for policy definition at the Application Binary Interface (ABI) level. We consider the roaming vulnerability found in OpenSSH, and demonstrate how using ELFbac would have provided strong mitigation with minimal program modification. This serves to illustrate the effectiveness of ELFbac as a means of privilege separation in further applications, and the intuitive, yet robust nature of our general approach to policy creation.

Modern applications often involve processing of sensitive information. However, the lack of privilege separation within the user space leaves sensitive application secret such as cryptographic keys just as unprotected as a "hello world" string. Cutting-edge hardware-supported security features are being introduced. However, the features are often vendor-specific or lack compatibility with older generations of the processors. The situation leaves developers with no portable solution to incorporate protection for the sensitive application component. We propose LOTRx86, a fundamental and portable approach for user-space privilege separation. Our approach creates a more privileged user execution layer called PrivUser by harnessing the underused intermediate privilege levels on the x86 architecture. The PrivUser memory space, a set of pages within process address space that are inaccessible to user mode, is a safe place for application secrets and routines that access them. We implement the LOTRx86 ABI that exports the privcall interface to users to invoke secret handling routines in PrivUser. This way, sensitive application operations that involve the secrets are performed in a strictly controlled manner. The memory access control in our architecture is privilege-based, accessing the protected application secret only requires a change in the privilege, eliminating the need for costly remote procedure calls or change in address space. We evaluated our platform by developing a proof-of-concept LOTRx86-enabled web server that employs our architecture to securely access its private key during an SSL connection. We conducted a set of experiments including a performance measurement on the PoC on both Intel and AMD PCs, and confirmed that LOTRx86 incurs only a limited performance overhead.

Failing to properly isolate components in the same address space has resulted in a substantial amount of vulnerabilities. Enforcing the least privilege principle for memory accesses can selectively isolate software components to restrict attack surface and prevent unintended cross-component memory corruption. However, the boundaries and interactions between software components are hard to reason about and existing approaches have failed to stop attackers from exploiting vulnerabilities caused by poor isolation. We present the secure memory views (SMV) model: a practical and efficient model for secure and selective memory isolation in monolithic multithreaded applications. SMV is a third generation privilege separation technique that offers explicit access control of memory and allows concurrent threads within the same process to partially share or fully isolate their memory space in a controlled and parallel manner following application requirements. An evaluation of our prototype in the Linux kernel (TCB textless 1,800 LOC) shows negligible runtime performance overhead in real-world applications including Cherokee web server (textless 0.69%), Apache httpd web server (textless 0.93%), and Mozilla Firefox web browser (textless 1.89%) with at most 12 LOC changes.