How to Compare Distributed Scale-Out File Storage Platforms for Use with Enterprise Workloads
This IDC Perspective discusses distributed scale-out file systems that meet a variety of enterprise file sharing requirements. Distributed scale-out file systems have been broadly used for a variety of different enterprise workloads. As IT organizations go through digital transformation and employ a variety of new applications, storage infrastructure requirements are changing. The lion's share of data that will be captured, stored, protected, and analyzed going forward will be unstructured, and scale-out file systems provide a good platform for many of these workloads. But there has been a lot of confusion in the market around these platforms, how they compete with other types of scale-out file systems (i.e., clustered scale-up and parallel scale-out), and how to evaluate which features are needed for which workloads. This document takes a detailed look at the functionality of distributed scale-out file system platforms, matching features and capabilities across different platforms with workload requirements. Customers considering a scale-out file system will be able to better select an unstructured storage solution that meets their needs after reading this document."Enterprise customers that do not differentiate between different scale-out file system architectures often run into unexpected limitations and challenges with file system workloads," said Eric Burgener, research vice president, Infrastructure Systems, Platforms and Technologies Group, IDC. "IT managers tasked with buying file system solutions need to understand the different capabilities of the three different scale-out file system options since that will help them select the best platform for their own needs."
Please Note: Extended description available upon request.
- Executive Snapshot
- Situation Overview
- Five Key Areas to Evaluate in Distributed Scale-Out File Systems
- Performance
- Determine the Latency You Require
- Is Your Workload Metadata Intensive?
- Consider Your IOPS Requirements
- Consider Throughput Requirements
- Will You Be Using Accelerated Compute?
- Availability and Resiliency
- How Much Resiliency Do You Require?
- What Snapshot Workflows Do You Anticipate Using?
- What Replication Capabilities Do You Need?
- Node, Rack, and Chassis Awareness
- What Sort of Failover Mechanisms Are Supported?
- Flexibility
- Do You Want Hardware Deployment Flexibility?
- How Much Granularity Do You Need in Applying Various Data Services?
- What Types of Access Methods Are Supported?
- What Type of Public Cloud Support Do You Need?
- Do You Plan to Use Automation with the System?
- What Deployment Flexibility Do You Want with the File System Itself?
- Ease of Use
- Does NFS Meet Your Performance Requirements?
- How Easy Is It to Configure the System to Simultaneously Meet Different Workload Requirements?
- Does the System Feature Any "Intelligent Data Placement" or Tiering Capabilities?
- How Comprehensive Are the Search Capabilities?
- Do You Need POSIX Compliance?
- Does the Vendor Offer Ways to Buy the System That Make It Easier to Deploy, Manage, and Support?
- Are There Additional "Services" You Would Otherwise Need to Supply That Are Integrated into the System?
- Does the Vendor Provide an "AIOps" Platform That Improves the Entire Customer Experience?
- Infrastructure Efficiency and Cost
- Does the Vendor Offer Software Innovations That Lower the Cost of Required Hardware to Meet a Given Performance Requirement?
- How Efficient Are the Vendor's Data Reduction Technologies?
- Does the System Support Extremely Low-Latency Technologies That Improve Server-Side CPU Utilization (and Hence Lower Server Costs)?
- How Efficient Is the Vendor's Cluster Interconnect Technology?
- How Does a System Perform Load Balancing?
- Does the System Support Mixed Media Types?
- Does the Vendor Offer an All-Inclusive Software Bundling Approach?
- How Good Is the Vendor's AIOps Platform at Driving Meaningful Business Value?
- Does the Vendor Offer a Range of Consumption Model Options?
- Advice for the Technology Buyer
- Learn More
- Related Research
- Synopsis