Last updated on: 2018-12-06
Authored by: Nick Shobe
Rackspace Proactive Detection & Response (PDR) on Google Cloud Platform™ (GCP) has two main components that need to be implemented in your GCP environment: the Network-based Intrusion Detection (NIDS) appliance infrastructure must be set up, and select vendor agents must be deployed.
To enable visibility of your GCP network, we deploy and maintain NIDS infrastructure in to a Rackspace PDR project that is peered to your other projects.
The Rackspace PDR and Rackspace GCP teams deploy and manage your GCP NIDS appliance infrastructure.
Our PDR teams deploy, manage, and monitor your NIDS appliances. Rackspace PDR has the following platform requirements:
If you have custom changes to your network environments, be sure to work with your GCP support team to implement them to comply with Rackspace PDR. Be sure that ingress and egress requirements pass through any security web application firewalls (WAFs) or gateway devices that sit in front of 0.0.0.0/0.
Rackspace PDR deployments require a separate project for the Rackspace PDR infrastructure. This project has its own networks that are isolated from the other projects and networks. Both of these networks use Amazon® Virtual Private Cloud® (VPC) Peering to provide connectivity with other projects.
The Rackspace PDR infrastructure project consists of, at a minimum, two NIDS appliances per GCP Region in use for high availability (HA).
Due to the General Data Protection Regulation (GDPR), European Union (EU) regions require a separate HA pair of NIDS appliances. Firewall rules ensure that connectivity between the EU and the United States is blocked for compliance with GDPR.
For more information, see Rackspace PDR Threat Manager network requirements.
Many appliactions terminate Secure Socket Layer (SSL) and Transport Layer Security (TLS) at the network edge with a load-balancer or web application firewall. If your application uses end-to-end encryption, see the Rackspace PDR SSL decryption guide.
Individual PDR agents are deployed and maintained by the Rackspace PDR team. However, we do have base requirements that must be met to ensure that our automated deployment system and PDR support team can access your instances to deploy or troublshoot agents and systems.
Following these steps helps to ensure successful agent deployments:
Due to the various vendors that we have selected to provide the nessessary telemetry to our systems, it is important that you select operating systems and kernel versions that are compatable with the vendor agents. For more information, see the Rackspace PDR System Requirements.
It is important that images taken from hosts that have Rackspace PDR agents deployed be prepared for deployment before using them as base images. Follow the Rackspace PDR imaging hosts guide to ensure Golden images are properly prepared.
Rackspace PDR on GCP uses a boot installation script to deploy and configure our vendor agents. We currently do not provide an alternative method for manually deploying agents on GCP, so proper configuration at boot is required to provision instances for Rackspace PDR.
The agents that are used to provide telementry to our Security Operations Center (SOC) have specific networking requirements that must be implemented. Use the Rackspace PDR Agent network requirements guide to correctly implement network ACLs and firewall rules for your platform.
For more information on the Threat Manager offering, see the Alert Logic upstream vendor documentation.
©2020 Rackspace US, Inc.
Except where otherwise noted, content on this site is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License