NooBaa V2 Addresses Next Generation of Cloud Storage Challenges

There’s a lot of buzz these days about supporting cloud migrations and having a multi-cloud strategy. Yet it’s an open question if those are valid, current concerns or a matter for some point in the future.
After all, migration away from Amazon Web Services (AWS) to Microsoft Azure or to Google Cloud Platform (GCP) is a mature-market concern in a world where users of those public cloud platforms still feel like innovators, or at least like early adopters. The bulk of workloads today may be modern in the sense that they’re virtualized, but a relatively low percentage of overall data and workloads are in the public cloud as of yet.
That said, AWS has been around since 2006, Microsoft Azure since 2010, and GCP since 2011. The time frame means it’s been more than long enough for innovators and early adopters to start getting swept up together in the frenzy of mergers and acquisitions of the last several years.
M&A is driving that early market for multi-cloud strategies and cloud migrations, according to Eran Tamir, Vice President of Product and one of the co-founders of NooBaa. Tamir briefed the ActualTech Media team recently as NooBaa was launching Version 2 of its hybrid storage service, and facilitating such migrations is a significant focus of the new release.
Version 2 follows up on the original, 2016 version of NooBaa. That original solution consists of a software-based control plane for aggregating all the storage in an environment, whether it’s in the public cloud or it’s on-premises as local storage. The aim is to provide a public cloud-like total cost of ownership in a locally managed service that has low CapEx, low OpEx, minimal setup and limited lifecycle costs. Part of the secret sauce is an innovative machine learning approach in the control software that optimizes for resilience, performance, economics and location when managing data placement.
With V2, the company is bringing a series of enhancements for organizations using NooBaa for cloud backup, infrastructure replacement, or repurposing of existing storage. A few of the new features include simple migration from AWS S3 API to Azure Blob and the other way from Azure Blob API to AWS S3; the ability to use lifecycle policies to automate data tiering and archiving; and faster installation with internal virtual storage.
One of the more intriguing capabilities of V2 involves a sophisticated multi-cloud scenario using a feature called Virtual Namespace. NooBaa’s Tamir describes the migration process from one cloud to another as a nightmare, partly due to egress fees. For a petabyte of AWS data, for example, it can add up to $100,000 in egress fees plus time and effort.
NooBaa V2 is available for a move from, for example, AWS to Azure. Banking on the notion that data is more often written than consumed, Virtual Namespace uses NooBaa’s control capabilities to take a staged approach.

AWS to Azure
AWS to Azure migration, Source:

In a case where a customer has an application feeding data to AWS S3, NooBaa builds a data container, called a NooBaa Namespace Bucket. By changing the configuration of the application, the application can continue reading data from AWS S3, while writing new data to an Azure Blob. In other words, there’s no need for a migration project. When enough new data is written to the Azure Blob, the customer can cut over to the new cloud without really moving any data and thus incurring massive egress penalties.
The capability is one of several reasons for organizations to give NooBaa a look, and NooBaa is trying to make that process as easy as possible. The company offers a free Community Version that supports up to 20 terabytes of data to allow organizations to thoroughly kick the tires before committing to this next-generation hybrid storage solution. I have personally deployed it in the ActualTech Media lab to spend some time playing with, and it’s quite simple to deploy and use. Screenshot of the configuration I deployed is below. I highly recommend trying it out for yourself!
My lab deployment of NooBaa’s storage fabric. Note that I was using EC2/EBS when I did this, but NooBaa now supports consuming S3 storage directly (and Azure Blob, etc).