Survey results show increased container use as part of phased cloud adoption
A survey of 250 IT organizations conducted by Denodo, a provider of a data virtualization platform, finds roughly 80% are leveraging some type of container technology, with Docker (46%) and Kubernetes (40%) being to two most widely employed.
More than a third of respondents (35%) also expressed a willingness to re-architect their applications to better fit the cloud architecture they are deployed on.
Overall, the survey finds more than three-quarters of respondents (78%) are running some type of application workload in the cloud. About 90% of those respondents are employing either Amazon Web Services (AWS) or Microsoft as their cloud service provider. The most widely deployed workloads involve analytics or business intelligence (BI) applications (66%), followed by logical data warehouse (43%) and data science (41%).
The survey also finds two-thirds of the data being accessed by workloads are in a structured format (68%) and that nearly half (47%) are making use of an object-based storage system.
Ravi Shankar, senior vice president and chief marketing officer for Denodo, says it’s clear applications are moving to the cloud in phases. In the first phase, applications were mainly monolithic applications deployed on virtual machines. The most recent wave of applications are based on cloud-native technologies that enable organizations to employ containers to build and update applications using microservices. However, Shankar notes that not every containerized application requires orchestration. Organizations are only employing platforms such as Kubernetes when containerized applications are complex.
The issue many IT organizations are now grappling with is the degree to which they may want to manage a platform such as Kubernetes themselves. Standing up a single Kubernetes cluster is comparatively trivial. Managing fleets of Kubernetes clusters requires a level of operational skill that most organizations lack today. As such, there’s a preference at least early on to rely on either a cloud service provider or a third-party IT services provider to manage Kubernetes clusters. Less clear is whether that represents a permanent shift in how IT infrastructure is managed or is simply a short-term option that will be relied on only until the internal IT team acquires sufficient Kubernetes expertise.
Regardless of the path chosen, the one thing that is clear is that as organizations consider their options in the wake of the COVID-19 pandemic, there will be a need for more applications that are both more flexible and resilient. Applications that can scale and up and down more easily in addition to being able to migrate between clouds will be highly prized. Many IT teams have already discovered, much to their chagrin during the pandemic, that existing monolithic applications do not lend themselves to being remotely accessed at scale to same level as a cloud-native application.
Of course, the economic downturn brought on by the pandemic might give some organizations a reason to pause their application modernization efforts. However, if those organizations expect to survive and thrive in the next era of IT, many of them will soon have a deeper understanding of the degree to which their legacy monolithic applications are holding them back.