Recommended Practices

Firewall

Abstract


Control Systems Cyber Security Defense in Depth Strategies

Research has shown that information infrastructures across many public and private domains share several common attributes in IT deployment and data communications for control systems. A majority of the systems use robust architectures to enhance business and reduce costs by increasing the integration of external, business, and control system networks. However, multi-network integration strategies often lead to vulnerabilities that greatly reduce the security of an organization, and can expose mission-critical control systems to cyber threats. This document provides guidance and direction for developing 'defense-in-depth' strategies for organizations that use control system networks while maintaining a multi-tier information architecture that requires:

  • Maintenance of various field devices, telemetry collection, and/or industrial-level process systems
  • Access to facilities via remote data link or modem
  • Public facing services for customer or corporate operations

Good Practice Guide on Firewall Deployment

In recent years, Supervisory Controls and Data Acquisition (SCADA), process control and industrial manufacturing systems have increasingly relied on commercial information technologies such as Ethernet, TCP/IP and Windows for both critical and non-critical communications. The use of these common protocols and operating systems has made the interfacing of industrial control equipment much easier, but there is now significantly less isolation from the outside world. Network security problems from the enterprise network (EN) and the world at large can be passed onto the SCADA and process control network (PCN), putting industrial production and human safety at risk.

Attack Methodology Analysis: SQL Injection Attacks

Database applications have become a core component in control systems and their associated record keeping utilities. Traditional security models attempt to secure systems by isolating core software components and concentrating security efforts against threats specific to those computers or software components. Database security within control systems follows these models by using generally independent systems that rely on one another for proper functionality. The high level of reliance between the two systems creates an expanded threat surface.

To understand the scope of a threat surface, all segments of the control system, with an emphasis on entry points, must be examined. The communication link between data and decision layers is the primary attack surface for SQL injection. This paper facilitates understanding what SQL injection is and why it is a significant threat to control system environments.

Backdoors and Holes in Network Perimeters A Case Study for Improving Your Control System Security

The Supervisory Control and Data Acquisition (SCADA) system of a natural gas utility was compromised resulting in a reduction of operation. The breach was discovered when operator interfaces became unresponsive and the system was no longer acquiring data. As a result, the system was disconnected from the network and a combination of manual operation overrides and limited fail-over to a backup server went into effect until the environment could be restored. Technicians troubleshooting the incident identified the deletion of several core application files on the primary control server as the source of the problem.

Common Control System Vulnerability

The Control Systems Security Program and other programs within the Idaho National Laboratory have discovered a vulnerability common to control systems in all sectors that allows an attacker to penetrate most control systems, spoof the operator, and gain full control of targeted system elements. This vulnerability has been identified on several systems that have been evaluated at INL, and in each case a 100% success rate of completing the attack paths that lead to full system compromise was observed. Since these systems are employed in multiple critical infrastructure sectors, this vulnerability is deemed common to control systems in all sectors.

Modern control systems architectures can be considered analogous to today’s information networks, and as such are usually approached by attackers using a common attack methodology to penetrate deeper and deeper into the network. This approach often is composed of several phases, including gaining access to the control network, reconnaissance, profiling of vulnerabilities, launching attacks, escalating privilege, maintaining access, and obscuring or removing information that indicates that an intruder was on the system. With irrefutable proof that an external attack can lead to a compromise of a computing resource on the organization’s business local area network (LAN), access to the control network is usually considered the first phase in the attack plan. Once the attacker gains access to the control network through direct connections and/or the business LAN, the second phase of reconnaissance begins with traffic analysis within the control domain. Thus, the communications between the workstations and the field device controllers can be monitored and evaluated, allowing an attacker to capture, analyze, and evaluate the commands sent among the control equipment. Through manipulation of the communication protocols of control systems (a process generally referred to as “reverse engineering”), an attacker can then map out the control system processes and functions. With the detailed knowledge of how the control data functions, as well as what computers and devices communicate using this data, the attacker can use a well known Man-in-the-Middle attack to perform malicious operations virtually undetected.

The control systems assessment teams have used this method to gather enough information about the system to craft an attack that intercepts and changes the information flow between the end devices (controllers) and the human machine interface (HMI and/or workstation). Using this attack, the cyber assessment team has been able to demonstrate complete manipulation of devices in control systems while simultaneously modifying the data flowing back to the operator’s console to give false information of the state of the system (known as “spoofing”). This is a very effective technique for a control system attack because it allows the attacker to manipulate the system and the operator’s situational awareness of the perceived system status. The three main elements of this attack technique are: 1) network reconnaissance and data gathering, 2) reverse engineering, and 3) the Man-in-the-Middle attack.

Security Implications of OPC, OLE, DCOM, and RPC in Control Systems

OPC is a collection of software programming standards and interfaces used in the process control industry. It is intended to provide open connectivity and vendor equipment interoperability. The use of OPC technology simplifies the development of control systems that integrate components from multiple vendors and support multiple control protocols. OPC-compliant products are available from most control system vendors, and are widely used in the process control industry.

OPC was originally known as OLE for Process Control; the first standards for OPC were based on underlying services in the Microsoft Windows computing environment. These underlying services (OLE [Object Linking and Embedding], DCOM [Distributed Component Object Model], and RPC [Remote Procedure Call]) have been the source of many severe security vulnerabilities. It is not feasible to automatically apply vendor patches and service packs to mitigate these vulnerabilities in a control systems environment. Control systems using the original OPC data access technology can thus inherit the vulnerabilities associated with these services.

Current OPC standardization efforts are moving away from the original focus on Microsoft protocols, with a distinct trend toward web-based protocols that are independent of any particular operating system. However, the installed base of OPC equipment consists mainly of legacy implementations of the OLE for Process Control protocols.