Test and Development Zone A Security Technical Implementation Guide

Up-to-date documentation is essential in assisting with the management, auditing, and security of the network infrastructure used to support the test and development environment. Network diagrams are important because they show the overall layout where devices are physically located within the network infrastructure. Diagrams also show the relationship and connectivity between devices where possible intrusive attacks could take place. Having up-to-date network diagrams will also help show what the security, traffic, and physical impact of adding a system will be on the network.Information Assurance ManagerInformation Assurance OfficerDCHW-1

Prior to connecting to a live operational network, such as the DISN, systems, at minimum, receive an IATO. A system without an IATO does not show adequate effort to meet IA controls and security requirements and may pose a risk to other computers or systems connecting to the operational network.Information Assurance ManagerInformation Assurance OfficerEBCR-1

An asset management system is used to send out notifications on vulnerabilities in commercial and military information infrastructures as they are discovered. If the organization's assets are not registered with an asset management system, administrators will not be notified of important vulnerabilities such as viruses, denial of service attacks, system weaknesses, back doors, and other potentially harmful situations. Additionally, there will be no way to enter, track, or resolve findings during a review.VIVM-1

It is important to restrict administrative access to the supporting network infrastructure and systems in the test and development environment, as it reduces the risk of data theft or interception from an attacker on the operational network.ECSC-1

An impersistent connection is any temporary connection needed to another test and development environment or DoD operational network where testing is not feasible. As any unvetted connection or device will create additional risk and compromise the entire environment, it is up to the Authorizing Official for the organization to accept the risk of an impersistent connection.EBCR-1, ECSD-1

To reduce the risk of compromise of DoD operational networks and data, application and system development needs to be limited to systems within a network segment designated for development only.ECSC-1

A HIDS or HIPS application is a secondary line of defense behind the antivirus. The application will monitor all ports and the dynamic state of a development system. If the application detects irregularities on the system, it will block incoming traffic that may potentially compromise the development system that can lead to a DoS or data theft.ECID-1, ECSC-1

A firewall provides a line of defense against malicious attacks. To be effective, it must be enabled and properly configured.ECSC-1

Change management is the formal review process that ensures that all changes made to a system or application receives formal review and approval. Change management reduces impacts from proposed changes that could possibly have interruptions to the services provided. Recording all changes for applications will be accomplished by a configuration management policy. The configuration management policy will capture the actual changes to software code and anything else affected by the change.DCII-1, DCPR-1

Without the approval of the Change Control Authority, data moved from the test and development network into an operational network could pose a risk of containing malicious code or cause other unintended consequences to live operational data. Data moving into operational networks from final stage preparation must always be vetted and approved.ECSC-1, ECSD-1, ECSD-2

Prior to release of the application receiving an IATO for deployment into a DoD operational network, the application will have a thorough code review. Along with the proper testing, the code review will specify flaws causing security, compatibility, or reliability concerns that may compromise the operational network.DCSQ-1, ECSC-1, ECSD-1, ECSD-2

Restricting access to source code and the application to authorized users will limit the risk of source code theft or other potential compromise.ECAN-1, ECCD-1, ECLP-1

If DoD production data is transferred to a test and development environment and personal or sensitive information has not been sanitized from the data, personal or sensitive information could be exposed or compromised.

Acting as the first hop into a test and development environment, the gateway can implement proper routing and provide a first layer of defense against attacks and other unintentional compromise or spillage of sensitive information into the operational network.DCSP-1, ECSC-1

In accordance with the DoD 8551.1 policy, the test and development environment may require external access to live operational data to perform final stage testing. All network connections for the test and development environment must make use of the PPS CAL at the appropriate boundaries.DCPP-1

To prevent malicious or accidental leakage of traffic, organizations must implement a deny-by-default security posture between test and development environments. All ingress and egress traffic not explicitly permitted between test and development environments must be denied. Such rule sets prevent many malicious exploits or accidental leakage by regulating the ports, protocols, or services necessary to each environment.ECSC-1

Remote access to the environment using unapproved encryption mechanism is inherently dangerous because anyone with a packet sniffer and access to the network can acquire the device's account and password information. With this intercepted information, a malicious user could gain access to the device, cause denial of service attacks, intercept sensitive information, or perform other destructive actions.EBRU-1, ECCT-1, ECCT-2

The VPN software on a host can be configured in either of two modes. It can be set to encrypt all IP traffic originating from that host, and send all of that traffic to the remote IP address of the network gateway. This configuration is called “tunnel-all” mode, because all IP traffic from the host must traverse the VPN tunnel to the remote system, where it will either be processed or further forwarded to additional IP addresses after decryption. Alternately, the VPN software can be set only to encrypt traffic that is specifically addressed to an IP at the other end of the VPN tunnel. All other IP traffic bypasses the VPN encryption and routing process, and is handled by the host as if the VPN relationship did not exist. This configuration is called “split-tunnel” mode, because the IP traffic from the host is split between encrypted packets sent across the VPN tunnel and unencrypted packets sent to all other external addresses. There are security and operational implications in the decision of whether to use split-tunnel or tunnel-all mode. Placing a host in tunnel-all mode makes it appear to the rest of the world as a node on the connected logical (VPN-connected) network. It no longer has an identity to the outside world based on the local physical network. In tunnel-all mode, all traffic between the remote host and any other host can be subject to inspection and processing by the security policy devices of the remote VPN-linked network. This improves the security aspects of the connected network, since it can enforce all security policies on the VPN-connected computer.ECSC-1

Attacks on virtual machines from other VMs through denial of service and other attacks potentially stealing sensitive data such as source code used in application development. It is imperative to keep DoD operational virtual machines on physically separate platforms from test and development virtual machines.ECSC-1

It is mandatory that data from an untrusted network or website that is to be used in a testing and development environment be downloaded through a secure perimeter. Bringing data directly from an untrusted network or downloaded from a personal computer or home Internet connection must be prohibited. Scanning data is crucial to ensure the integrity of the information prior to deployment for T&D processes. While not an all-inclusive list, data in this situation includes OS patches, application updates, operating systems, development tools, and test data. In the T&D environment, there will typically be one or more IA-compliant systems accessing a secure Internet connection. If a secure Internet connection is not available, such as in Zone D, a connection in another zone can be used and the data moved by approved physical media into the zone. Scanning the data with an anti-virus program will reduce the risk of exploits and of having vulnerable systems in the T&D environment taken over. Downloading data from a single workstation for all zone environments is acceptable. Organizations with NIPRNet connections must download all data through their NIPR connection for scanning at the IAPs. Contractors or other DoD organizations without any direct NIPRNet connectivity will need to use a secure Internet connection following all applicable DoD IA policy and STIG requirements.

Without policies and procedures in place, the organization will not have the authority to hold personnel accountable for improperly handling or transporting data into the test and development environment. The documents need to include guidance for both physical and electronic data migration.