Virtualization and Cloud executives share their predictions for 2016. Read them in this 8th Annual VMblog.com series exclusive.
Contributed by Avinash Lakshman, CEO and Founder of Hedvig
Storage Technology and Data Center Predictions for 2016
Here at Hedvig we see a tremendous amount of innovation happening in the data center. We’re seeing new hardware like ARM gain steam; we see a wholesale shift to Docker and OpenStack for more agile clouds; we see companies moving beyond just tinkering with Hadoop; and we see companies building parallel, software-defined infrastructure to modernize their data centers – a phenomenon Gartner calls Bimodal IT.
What do these trends have in common? The answer: use of emerging technology in production data centers. All of these technologies have matured to the point where mainstream enterprises are comfortable architecting them into the foundation of their data centers.
With that in mind, here are five production-ready predictions Hedvig is making for 2016:
1. ARM will climb to 5% of the server chip market.
ARM servers make up roughly 1% of the market. We predict that will climb modestly to 5% next year. Although that seems insignificant, it has huge downstream ramifications on data center cost and efficiency. The chip design firm, ARM Holdings PLC, predicts ARM share will grow to 20% by 2020, citing an average cost of $100-200/chip versus Intel’s $600/chip as a big reason. We think the jump to 5% next year will be fueled by scale-out, software-defined infrastructure that needs to run on commodity servers. Today’s cloud and enterprise data centers will seek smaller, more efficient servers to prevent infrastructure sprawl. As a result, we’ll see our first ARM-based software-defined storage deployment by the end of the year.
2. Production adoption of containers will plateau at 50%.
DevOps.com surveyed 285 organizations and found 38% are already using containers in production environments. The same survey found 65% of respondents plan to deploy containers in production in 2016. We feel that’s a bit optimistic, and think production adoption will climb moderately to around 50%. Why the difference? Bottlenecks. Infrastructure components like security, stateful data storage and monitoring are still not where they need to be fore widespread production adoption. Although the broader ecosystem is knocking down these issues, it will take 12-18 months for best practices and reference architectures to permeate the community.
3. In contrast, only 30% of companies will deploy OpenStack in production.
Containers and OpenStack are not mutually exclusive. In fact, we believe strongly that the data center of the future will be a mix of VMs and containers, with OpenStack as an orchestration layer atop both. But OpenStack is not growing as quickly as containers-Docker in particular. A 2015 Red Hat survey found 16% of organizations are using OpenStack in production. There’s good momentum as companies get serious about OpenStack clouds and we previously predicted the number of production deployments will roughly double. Large enterprises will lead the way; they have the most resources to allocate to the platform.
4. Enterprise Hadoop becomes a “virtual-first” deployment.
Several studies, including a great benchmark from VMware, found that virtualizing Hadoop actually improves performance. Surprising, but true. It’s also the cherry on top. The real reason is similar to why enterprises have always virtualized apps: better hardware utilization, better DR and operational savings. In 2016, companies will take Hadoop from tire-kicking and departmental experiments to production by adopting a “virtual-first” stance. In doing so, Hadoop can be more effectively run on shared storage (studies have shown this introduces 8% or less performance decrease). Why? To get better economics for a central data lake that leverages snapshots, clones, deduplication, and storage tiering – all needed to make Hadoop production-ready.
5. We’ll see the first multi-datacenter software-defined storage deployment.
Software-defined storage (SDS) is quickly becoming a production-ready technology. Large organizations realize that decoupling storage functionality and running it as API-driven software on commodity infrastructure not only reduces costs by 60% or more, but also provides a more flexible infrastructure to accommodate ever-changing business needs. However, to date, production SDS has been limited to single site deployments except for some object storage used for data archiving. In 2016 we predict enterprises will not only use SDS for a wider swath of production applications – including databases, Hadoop, Exchange, etc. – they’ll do so with replication that automatically protects those production apps across four or more data centers.
##
About the Author
Avinash Lakshman is the CEO & Founder of Hedvig. Before starting Hedvig, Avinash built two large distributed systems: Amazon Dynamo and Apache Cassandra. As the pioneer of NoSQL systems, Avinash is passionate about using distributed systems to disrupt a storage space that hasn’t seen any real innovation over the last decade.
