Research Library

Quantum ActiveScale – Product Brief

Last updated October 5th, 2021 for Cloud Storage offering. Concise 8-page overview of Quantum ActiveScale. Covers highlights, potential, usage, & strengths/weaknesses. Download the full Product Brief now!

ActiveScale is a scale-out object storage system acquired by Quantum in March 2020. Previously, ActiveScale was sold by Western Digital after their purchase of Amplidata in March of 2015. Focused on the storage demands of big data, media & entertainment file data, archives, and content repositories, the ActiveScale object storage system has the ability to scale to very large capacities with large numbers of objects while providing the availability, durability, and performance needed for those use cases.

There are three current models in the ActiveScale family, the P100, P100E3 and the X200. All systems execute the same embedded  software.

Using commodity based platforms; ActiveScale object storage systems are built with a predefined configuration using dedicated 10GigE or 25GigE for the X200. A single namespace or storage pool is maintained across the multiple nodes.

Forward Error Correction with rateless erasure coding is used to distribute data across nodes and across geographies. Object data can be accessed in parallel, increasing the performance for
bandwidth of data transfer. The controllers manage the spreading function with rateless erasure coding and control access. The distribution function will balance capacity across nodes, managing the non-disruptive addition of storage nodes. GeoSpread uses the same rateless erasure coding to distribute data across geographically dispersed systems. Agents are installed on nodes for out of band management and do continuous background integrity checking and healing. ActiveScale 5.5 added the ability for users to choose between strong and eventual consistency using Asynchronous GeoSpread.

Object caching in SSD or HDD caches is used for performance acceleration and optimizes transfers for geographically dispersed data or high-latency networks. Read caching can be enabled on any number of controllers and objects are only cached in a single controller. Writes are write-through and cached data is invalidated on a write.

Quantum ActiveScale Product Brief includes:

  • Introduction
  • Highlights
  • Usage
  • EvaluScale Product Review Methodology
  • Evaluator Group Opinion
Available with Subscription or Purchase. Contact us below.

Forgot your password? Reset it here.