In this BlogBytes article, we will look up, scan the sky, and examine the clouds. What are cloud data services? What is cloud repatriation? Why repatriate cloud data? And, a whole lot more about moving data in and out of the cloud. This is the first of a two-part BlogBytes series about cloud and on-premises data storage – Let’s get cloud watching.
WHAT ARE CLOUD DATA SERVICES?
You probably know this, but to stay on the same wavelength, cloud services are computing services provided by a third party over the internet, giving users access to computing and storage resources.
WHAT IS CLOUD REPATRIATION?
It’s the practice of moving data, workloads, or even applications from a cloud provider back to on-premises data centers. Another way to describe it is an organization that migrated data and systems to the cloud decides to bring some or all of them back to on-site local systems.
WHY REPATRIATION FROM THE CLOUD?
There are numerous reasons to repatriate data from the cloud back to on-premises environments. As described at techradar.com, “businesses are facing a mix of pressures that are forcing them to reconsider their reliance on public cloud infrastructure. Surging cloud costs, mounting regulatory demands (both local and international), issues around data sovereignty, and geopolitical concerns regarding dependency on a few U.S.-based cloud providers are all prompting organizations to explore alternatives.” The article states that, “A 2024 CIO survey from Barclays revealed that 83% of enterprises are planning to move workloads from public cloud to private or on-premises solutions, up from just 43% in 2021.”
WHAT ABOUT: COST – PERFORMANCE – SECURITY – AI IMPACT?
As we just learned, cloud repatriation is occurring because organizations are getting savvier about where they want to keep their data.
There are numerous reasons why organizations are repatriating data back to on-premises storage, but cost, performance, security, and the impact on AI are high on the list. Let’s look closer at the details. A hyscaler.com article explains that the reasons many organizations are repatriating data include:
- Cost Optimization and Budget Control: data egress fees can be substantial and unpredictable, while constant API calls and storage access charges accumulate rapidly.
- Performance and Latency Requirements: applications requiring real-time processing often perform better on dedicated infrastructure closer to end users.
- Security and Compliance Concerns: certain industries face stringent regulatory requirements that make private infrastructure more attractive.
- The AI Revolution’s Impact: AI workloads have fundamentally challenged traditional cloud economics because AI can necessitate massive computing power requirements and needs fine-tuned hardware optimization.
When considering AI workloads, the amount of data and compute power needed can vary by project, but many models can require ‘data lakes’ with images, audio, video, and more, reaching petabytes of information. Accessing these large workloads could best be managed with local systems that can move data from tape systems to high-speed processing and back to tape for low-cost, long-term, safe storage.
In the next BlogBytes of this series, we’ll delve further into the hidden economics of cloud data repatriation, which will include an examination of the total cost of ownership for cloud storage and onsite tape storage, and lots more. Stay tuned!
