Flexibility, multi-clouds and the edge -- storage and data predictions for 2022
Thanks to the IoT and the collection of customer data online, businesses are accumulating more information than ever before.
This means that the need for storage isn't going to go away -- even though it may now be 'out of sight, out of mind' in the cloud. So, what do industry experts think is the future for storage and data as we head into 2022?
Surya Varanasi, CTO of StorCentric, thinks the ability to support different workloads will be important. "In 2022, organizations will seek a data storage foundation that enables them to support a mix of workloads. The storage will allow for flexible configurations and simplified expansion to meet a wide variety of capacity and performance requirements. Next, organizations will seek a solution that provides multi-protocol support. Certainly, at a minimum, the storage will support block (iSCSI, FC) and file (NFS, CIFS/SMB). However, given the increasing desire to run cloud-native applications, backup and restore critical data, as well as archive data in the cloud, S3 object storage support will become a must-have as well."
Luke Han, CEO of Kyligence, believes businesses will hang on to Hadoop for storage even as the platform declines:
In 2022, we can expect the continued decline of the Hadoop platform, even though like some tough weeds in your garden the roots and trailers of Hadoop will be hard to completely eradicate. Expect CIOs and data teams to continue to de-emphasize Hadoop and to continue the process of removing it from their production data stack.
Also look for IT departments to continue to make their on prem implementations look and function like the public cloud. In the near term, organizations may continue to use the Hadoop File System (HDFS) as a storage platform until a better private cloud storage solution can be devised.
Steve Touw, CTO of Immuta thinks the move to the cloud will continue. "Customers are increasingly using multiple cloud providers across their IT infrastructures. Organizations are well underway in their shift to the cloud for cloud computing for data processing and storage: in fact, 81 percent project they will become 100 percent cloud-based or primarily cloud-based in the next 12-24 months. Many organizations fear being forced to lock-in to a single cloud vendor platform. Multi-cloud architecture means companies can forgo the 'lock-in' and instead operate across all cloud platforms. This trend is not going away, but rather accelerating, and cloud vendors must get on board as it adds a layer of backup and recovery, scalability, and improved security."
Karthik Ranganathan, co-founder and CTO at Yugabyte, believes organizations will need more flexible databases. "Monolithic RDBMSs were not designed to meet the needs of cloud native applications. The rise of microservices, cloud infrastructure, and DevOps puts pressure on traditional systems of record. Companies are increasingly seeking databases that can run anywhere that cloud native applications are deployed; across private, public, hybrid, and multi-cloud environments. To satisfy demand, databases need to combine powerful RDBMS capabilities with cloud native resilience, scale, and geo-distribution. They also need to quickly, easily, and non-disruptively scale to handle peak demand."
Hao Zhong, co-founder and CEO of ScaleFlux, foresees the need for local storage at the edge. "There will be an increase in demand for edge computing. This increase is being fed by the need to power a growing number of base stations, the fixed transceivers that serve as the main communication point for mobile technology as 5G (and even 6G) grow in popularity. Also, as the number of self-driving cars increases, so does the need for local storage, as well as growth in remote processing power in the car itself. Edge will lead infrastructure and even cloud growth in the coming year."
Anand Babu Periasamy, co-founder and CEO at MinIO, agrees on the growth of edge storage, "While the edge has two primary topologies, edge cache and edge storage, the growth will come in the latter. Edge storage increasingly demands containerization because it does not look like a mini data-center but rather a distributed system of endpoints."
Periasamy also sees software defined storage giving way to container defined storage:
Kubernetes has already won -- this much we know. Since Kubernetes only knows how to manage containers, not bare metal or appliances, this is where the world migrates to. Interestingly, this extends to services too. Only when everything is software defined can virtualization be optimized. Only when everything is containerized can Kubernetes be optimized.
This makes container-defined everything the trend to watch in 2022. For those organizations struggling to make the transition from appliances to software, the bridge to container defined may be too far.
Morgan Littlewood, senior vice president of product at iXsystems, thinks, "The flexibility of scale-out storage will underpin a growing number of storage deployments in 2022. First proven in larger environments, small and medium size enterprises have proven choices in scale-out storage that are both affordable and easier to manage. It’s important that file or object workloads can be easily supported together as the needs of applications often change more quickly than the infrastructure."
Gary Ogasawara, CTO of Cloudian, thinks object storage will accelerate high performance computing (HPC) innovation. "As HPC deployments have become highly distributed and begun to exceed exabyte scale, it’s become clear that the storage component of HPC infrastructure needs greater focus. To continue making advances in supercomputing, organizations will require highly scalable, software-defined storage that can accommodate massive data sets while easily leveraging any hardware innovations on the computing side. Parallel file storage alone cannot provide this scalability and flexibility. As a result, more organizations will use object storage as the primary storage for supercomputing deployments."
Alluxio's founder and CEO Haoyuan (H.Y.) Li sees the future in data lakes. "Data lakes are rising to prominence and structured data is transitioning to new formats. In 2022, open-source projects like Apache Iceberg or Apache Hudi will replace more traditional Hive warehouses in cloud-native environments, enabling Presto and Spark workloads running more efficiently on a large scale."