Skip to main content

Virtual MachinesGoogle Cloud Platform Setup Guide

Table of Contents

Request Images to Deploy in GCP

To request access to the GCP images, visit the Virtual Images section in the Exabeam Community. You will need to provide the following information:

  • The name of the member of your organization that will perform the deployment in GCP

  • The business email address of that member of your organization

Your credentials will be added to the Exabeam Google Group to access product images.

Note

Please review all specifications for your platform and ensure you have sufficient resources to deploy Exabeam images. Additionally, please ensure you have valid Exabeam licenses for the product(s) you will implement.

Exabeam Platform Specification Table for Virtual Platforms

These are the minimum operating specifications needed to run your Exabeam product. We do not support hybrid deployments (cross environment deployments). All nodes must be in the same subnet.

Be aware that vCPU is not the same as the number of CPUs or Cores for the processor. A vCPU is typically equal to the number of threads in the processor.

The tables below details the CPU and memory allocation required for Exabeam products to operate optimally, with the following provisioned:

  • The host is not shared with any other product or resource.

  • All virtual appliances must be approved by Exabeam prior to deployment.

  • Storage requirements:

    • Use local drives or an NAS or SAN with a 10Gbit/s link (iSCSI or Fibre Channel) and block storage access; the underlying drives must also meet the SSD/HDD IO performance requirements.

      Important

      Configuration and maintenance of NAS/SAN is the responsibility of the customer. Exabeam cannot debug issues related to misconfiguration or performance with NAS/SAN.

      Important

      NFS is not supported.

    • Do not use VMFS with extents spread across multiple drives and managed by ESX.

    • To avoid performance issues, SSD drive files MUST be put on an SSD backed data store.

      Note

      To ensure adequate virtual drive capacity, a 10% additional storage buffer is required on the physical drives. For example, when an OVA requires 215 GB of virtual space, the appliance requires a physical drive with 236.5 GB of space.

    • Disk storage must be configured with thick provisioning (eager zeroed recommended). Snapshots are not supported because they switch the environment from thick to thin provisioning.

  • Network requirements:

    • Network card must support paravirtualization (VMXNet v.3 adapters).

Advanced Analytics Node Type

GCP

Physical Master Node including Hypervisor

Virtual Hypervisor

4 vCPU

Memory: 10 GB

1 x 32 GB (SSD)

OVA Virtual Master Node

vCPU: 40 Cores

Memory: 256GB

Storage:

  • 1 x 150 GiB (SSD)

  • 3 x 894 GiB (SSD)

  • 6 x 1860 GiB (HDD)

Virtual Hypervisor

4 vCPU

Memory: 10 GB

1 x 32 GB (SSD)

OVA Virtual Worker Node

vCPU: 20

Memory: 128 GB

Storage:

  • 1 x 150 GiB (SSD)

  • 3 x 894 GiB (SSD)

  • 6 x 1860 GiB (HDD)

Table 1. Advanced Analytics Node Specifications


Node Type in Exabeam Cluster

GCP

Virtual Hypervisor

OVA Virtual Master or

Worker Node

vCPU: 20

Memory: 192 GB

Storage:

  • 1 x 240 GiB (SSD)

  • 2 x 1920 GiB (SSD)

  • 9 x 4000 GiB (HDD)

Table 2. Data Lake Node Specifications


For clusters with 21 nodes or more, an additional three management nodes are required for cluster management operations, health monitoring and other critical functions.