GCP Deployment Specifications

On This Page

Overview

This document lists the hardware specifications for deployment of version 3.5.4 of the MLOps Platform ("the platform") on Google Cloud, also known as Google Cloud Platform (GCP); for details, refer to the Google Cloud Compute Engine documentation.

Docker Registry Requirements
The platform is deployed with a default, on-cluster docker registry. This registry can only be used for playground environments, as it is volatile and images are lost when it is restarted. Operational clusters must be connected to an external Docker registry. Create the registry in your cloud, then configure it either during installation (See Custom User Docker Registry in installation), or post-installation through the platform dashboard.
Note
All capacity calculations in the hardware specifications are performed using the base-10 (decimal) number system. For example, 1 TB = 1,000,000,000,000 bytes.
Warning

  • Provisioning of the servers is handled automatically by the platform installer (Provazio).
    Don't attempt to provision the servers manually prior to the deployment.

  • The data-node instances include Non-Volatile Memory Express (NVMe) SSD-based instance storage, which is optimized for low latency, very high random I/O performance, and high sequential read throughput. The data doesn't persist on the NVMe if the instance is stopped.
    Don't attempt to shut down any of the data nodes, as it will erase the data.

Hardware Configurations

The platform is available in two configurations, which differ in a variety of aspects, including the performance capacity, footprint, storage size, and scale capabilities:

Iguazio Playground

A single data-node and single application-node cluster implementation. This configuration is designed mainly for evaluation trials and doesn't include high availability (HA) or performance testing.

Note
You can deploy a Proof of Concept (POC) in the Iguazio playground configuration. Be aware that the POC cannot be used as development environment or for a production environment.
Operational Cluster
A scalable cluster implementation that is composed of multiple data and application nodes. This configuration was designed to achieve superior performance that enables real-time execution of analytics, machine-learning (ML), and artificial-intelligence (AI) applications in a production pipeline. The minimum requirement for HA support is three data nodes and three application nodes.
GCP Production Cluster

GCP Data-Node Specifications

Data nodes in platform GCP deployments must fulfill the following hardware specification requirements:

Component Specification
Instance type n2-highmem-16
vCPUs 16
Memory 128 GB
Data disks (local storage) 8 x 375 GB NVMe SSD
OS boot disk Premium SSD; 400 GB (minimum)
Usable storage capacity 1 node (Playground) — 2 TB;
3 nodes (Operational Cluster) — 3.5 TB

GCP Application-Node Specifications

Application nodes in platform GCP deployments are supported only on Google Kubernetes Engine (GKE) and must use one of the following instance types. Choose the type that best fits your requirements. For specification details for each type, refer to the GCP documentation.

Note
All of the supported application-node configurations also require a 250 GB (minimum) premium-SSD OS boot disk.
CPU-Based Instances
  • c2-standard-16 (default configuration)
  • c2-standard-30
  • c2-standard-60
GPU-Based Instances
  • n1-standard-16
  • n1-standard-32
  • n1-standard-64
  • n1-standard-96

Backing up the Platform

Caution
To ensure safety of your MLOps Platform data, you must periodically back up your data and configuration.

Allocate one Filestore in your GKE for backing up your platform. See full details in Backing Up the Platform.

See Also