2019 Data Storage Predictions

Data Storage
Insights by Scott Sinclair, ESG Senior Analyst

Some clearly defined expectations from 2019 are:

Public cloud adoption will continue, even accelerate.

All-flash adoption will continue with increased deployments v

Business success will become even more dependent upon IT services and capabilities, and technologies like analytics, AI, and IoT are expected to grow substantially.

But the predictions are based on these for 2019:

Back to basics: More companies will bring a workload back from the public cloud to on-premise:  Last year, 41% of businesses brought at least one workload back from public cloud infrastructure, to be run on premises. These moves are not because there is a problem with the cloud, but because of a lack of due diligence before adoption- such as lack of understanding in the suitability of the workload or not understanding the specific performance or sensitivity requirements. As the excitement on cloud adoption increases, due diligence may take an even more back seat, and lead to even more failures, causing more moves back to on-premises. With time as organizations become more adept at workload analysis and modifications required before the cloud becomes their choice, these will be ironed out. But there is time for that.

In 2019 the NVMe over Fibre Channel adoption growth is expected to outpace NVME over Ethernet.  In addition, many of the ecosystem support components, such as OS support, hypervisor support, and multi-pathing, will have to be resolved for NVMe over FC shortly.

Trending tech may miss goals: Despite the hype, a significant percentage of analytics, AI, and IoT projects will miss their goals. This will be due to lack of adequate Metadata handling. AI requires big volumes of data, and as this volume increases, the efficiency of the AI tool increases with it. For years enterprises have been storing data, but with no particular structure, and the result is disaggregated data that is difficult to analyze. This will defeat the purpose of AI.  While many attempts have been made to solve this problem through Metadata handling, especially for bigger environments, the real solution is still elusive. For AI to succeed, getting to solve this problem will become a priority in 2019.

Previous articleTrends to Watch Out For In 2019: Analyst Thoughts
Next article2019 Predictions for Cloud Services & Orchestration and DevOps