Blog

Subscribe to Email Updates

4 Requirements of Modern Backup and Archive

by Catherine Chiang – February 20, 2018

As enterprise datasets grow at unprecedented rates, with the majority of it being unstructured data, requirements for modern backup and archive have expanded beyond the capabilities of legacy secondary storage systems.

Modern backup and archive must have the following four requirements to meet the demands of massive unstructured data:

Policy-Based Data Management

Policy-based workflows for backup and archive streamline data management by allowing administrators to easily set policies for automatic backup and tiering to cloud.

Another big problem with massive datasets is not knowing what’s there, so it’s incredibly useful to have a single consolidated tier with indexing and search. Autodiscovery of shares and exports helps administrators easily discover what data needs to be protected.

Data Movement Engine

Moving data becomes extremely difficult when datasets are large. Legacy backup solutions use single-threaded protocols to move data, but this fails for petabyte-scale data.

A modern backup and archive solution must use highly parallel streams to move data. Another requirement of modern data movement is latency awareness, which enables backups to run continuously without creating backup windows.

Cloud-Native Services

As data grows quickly both on-premises and in cloud, enterprises face the challenge of data management in a hybrid world. Cloud-native architecture for on-premises infrastructure helps to bring cloud benefits to enterprise datacenters.

These cloud benefits include scale-out architecture, resiliency, and agility. Scale-out architecture allows enterprises to scale their solution without creating silos, which becomes essential as data grows large. The distributed nature of cloud-native architecture results in resiliency, protecting against potential failures. Lastly, users experience agility due to nondisruptive software updates.

As-a-Service

Finally, as the demands for managing massive unstructured data grow, as-a-Service providers offer a scalable, more agile alternative to legacy systems.

“as-a-Service” means that instead of buying hardware and then dealing with the day-to-day maintenance and troubleshooting of the secondary storage solution, customers pay for an all-inclusive service which alleviates management overhead.

as-a-Service offerings take care of monitoring, diagnostics, failure management, and software updates. This is particularly valuable as data grows larger because it actually ends up saving enterprises money by lowering total cost of ownership (TCO). And unlike traditional managed services, which employed customer service teams that get expensive as data grows, as-a-Service vendors use software to provide efficient and cost-effective services.

Related Content

PAIGE and Igneous Build Industry-Leading Compute Cluster for Healthcare AI

January 16, 2019

PAIGE’s mission to revolutionize the diagnosis and treatment of cancer through machine learning requires an extremely large dataset of high resolution slide images. To do so, they are building an industry-leading compute cluster for healthcare AI. The team needed to not only protect all of this unstructured data, but also programmatically move and process small subsets of the overall dataset on demand for high performance computations.

read more

Coming Soon: A New Approach to Protecting Datasets

December 17, 2018

Unstructured data has grown at an annual compounded rate of 25% for the past ten years, and shows no sign of slowing. For most organizations, “data management” for unstructured data has really just meant capacity management, i.e. increase capacity to keep up with data growth. This model worked at moderate scales, but as datasets have increased in size, complexity, and quantity, it has pushed the scales into petabytes of data with billions of files, and overwhelmed budgets. Enterprises are now asking for data management strategies that do more than just provide continuously increasing capacity.

read more

Accelerating Image Analysis and Cancer Diagnosis with AIRI from Pure Storage and Igneous

November 28, 2018

Artificial Intelligence (AI) has various applications today, from self-driving vehicles to optimizing workflows in manufacturing operations to detecting malware on the internet. Deep learning is a form of AI where multi-layer neural networks are utilized to transform input data into progressively more defined and useful outputs. Deep learning differs from machine learning (ML) in that ML focuses on the development of task-specific algorithms that can be applied to specific problems, while deep learning focuses on extracting information at multiple levels.

read more

Comments