A distributed cloud lets organizations manage multi-cloud and edge deployments, as well as traditional data center deployments, as one simple cloud.
The increasing adoption of the Internet of Things (IoT), artificial intelligence (AI), and machine learning is changing the distribution of apps and data, which are rapidly being deployed at multi-cloud and edge environments. By 2022, Gartner expects more than 50 percent of enterprise data will be created and analyzed at locations like multiple clouds and the edge rather than centralized data centers. Multi-cloud and edge deployments are supporting critical new use cases, but organizations are currently facing considerable problems in achieving the transition to these highly distributed environments.
See also: The Benefits of Using Self-Service, Elastic, Cloud Infrastructures for Continuous Intelligence
A recent global survey of more than 400 IT decision-makers shines light on the issues facing organizations as they look to support apps and data across multiple clouds and edge environments, while also revealing more about what organizations specifically look to gain from these deployments. Let’s start with multi-cloud.
Multi-cloud: Better availability and reliability, but security, connectivity, and operational challenges are prominent
According to the survey, the vast majority of organizations are currently pursuing multi-cloud deployments, with 97 percent of respondents saying that they are planning to distribute workloads across at least two clouds. The top driver for multi-cloud adoption is a need to improve availability and reliability for their applications (63 percent). Beyond that, multi-cloud serves the purpose of satisfying regulatory and compliance requirements (47 percent) and enables enterprises to leverage the best-of-breed services from each provider (42 percent).
Here’s how multi-cloud improves availability and reliability: if one cloud platform goes down for some reason, any apps will still be available to use in at least one more cloud. When it comes to regulatory and compliance concerns, a multi-cloud approach makes it possible to put an app’s data in a certain geography in case local law mandates it (which is common for highly sensitive data, such as financial information). Lastly, using multiple clouds gives enterprises the opportunity to take advantage of the specific strengths of every cloud. For example, organizations can take advantage of Google Cloud Platform’s excellent machine learning capabilities or leverage Microsoft Azure for its robust support of Office 365 databases.
But organizations face major security and connectivity challenges in supporting multi-cloud deployments because of the varying functionality and user experience of each cloud platform. Managing workloads across different clouds also comes with significant operational challenges. According to the survey, achieving reliable and secure connectivity between providers (60 percent), navigating the different consulting and support processes between the clouds (54 percent), and dealing with different platform services (53 percent) rank among the top issue associated with managing multi-cloud workloads. s
In drilling down specifically on the problem of connecting between clouds for a shared workload, the survey showed the biggest challenges were security (54 percent), reliability (44 percent), and performance (39 percent). All the challenges reflect the wider issues of cloud siloing and cloud lock-in. There’s simply not enough interoperability between the major clouds, and it’s too difficult to move data or spread apps across these platforms. This naturally results in confusion, security lapses, performance issues, and other problems. The big cloud platforms will need to accept that major enterprises are adopting multi-cloud approaches and introduce functionality to support these deployments.
Edge: IoT and the need to analyze data locally drive deployments, but organizations struggle to meet infrastructure demands and cohesively manage disparate edge apps
Unsurprisingly, the survey revealed that IoT (57 percent) was the top use case, followed by smart manufacturing (52 percent) and content delivery (46 percent). When asked why they were putting these workloads at the edge instead of public or private clouds, respondents cited the need to control and analyze data for such use cases locally (54 percent) and pointed out that there’s too much latency when sending edge data to public cloud-based apps (47 percent). This reflects the basic value proposition of edge computing – the need to process data where it’s created due to the latency and cost of sending it to the cloud and back. A self-driving car, for example, needs to make decisions in milliseconds, not seconds, so the data must be analyzed at the edge rather than the cloud.
But edge deployments face major infrastructure challenges, as well as headaches in managing apps spread across various edge sites. Respondents indicated the biggest business concerns in putting apps at the edge are the difficulty in managing apps across multiple edge locations (44 percent) and an inability to accommodate the IT infrastructure needed to host and operate at the edge (38 percent). Similar to multi-cloud deployments, organizations lack a way to cohesively manage apps spread across edge sites. As for infrastructure, the market has not yet produced solutions that provide enough processing power in a small enough form factor as these use cases demand.
The survey also looked at the more specific technical challenges organizations face in supporting edge deployments. Again, respondents highlighted management and infrastructure issues: The top challenge is integrating cloud-native workflows like automation, CI/CD, and performance management (69 percent), followed by trouble installing a full set of application infrastructure (compute/storage/network/security) (67 percent).
A new way forward
Multi-cloud and edge deployments offer massive value to organizations, but several common challenges are preventing these organizations from maximizing this value. Multi-cloud is vexed by security, connectivity, and operational problems. The major cloud platforms lack interoperability and visibility between one another, preventing organizations from establishing common policies or a consistent operational experience for multi-cloud deployments. When it comes to the edge, the top issues are managing apps across disparate edge sites and a lack of adequate infrastructure.
Cloud platforms will add new functionality and edge infrastructure vendors will make progress to help with these issues, but there’s only one option that will completely overcome them: the distributed cloud, which was recently named by Gartner as one of the top ten strategic technology trends for 2020. The distributed cloud is an emerging approach that allows organizations to manage multi-cloud and edge deployments, as well as traditional data center deployments, as one simple cloud. This means being able to manage, operate, and secure all these highly distributed deployments as a single cloud, eliminating the pain points associated with multi-cloud and edge deployments uncovered by the survey. The distributed cloud will take a few years to fully materialize, but it will eventually catch on between 2022 and 2024.