What is distributed cloud and how does it work?
Distributed cloud is public cloud -- any third-party cloud computing service offered over public internet -- deployed in more than one location, ideally in proximity to whoever or whatever wants to access it.
These locations include but aren't limited to within the public cloud provider's infrastructure, on-premises or even in another cloud provider's data center. This flexibility is one of the benefits of distributed cloud, which further distinguishes itself by offering users a single point of control despite its multi-cloud, multi-location architecture.
A robust public cloud infrastructure lays the groundwork for edge computing, decreasing the physical distance between cloud computing services, such as storage, and the end user or application requesting the service.
Conceptually, distributed cloud also borrows aspects from distributed computing, where application components are spread across different machines, which communicate with one another to provide the requested cloud service. However, with distributed cloud, the application isn't broken up when deployed. Instead, the entire technology stack is installed wherever the customer wants.
The service's centralized cloud is thus extended across physical locations, creating a network of nodes or "micro-cloud" satellites. The client, though, accesses the entire cloud service as a seamless experience managed from a centralized interface. Meanwhile, the cloud service provider maintains control of the entire distributed infrastructure from a single point, improving maintenance and logistical outcomes, such as updates.
Key benefits of distributed cloud
Distributed cloud contributes to an invested client's operational success in a wide array of business areas:
- Lower latency. By shrinking the distance between data and the party attempting to access it, performance is improved through decreased latency.
- Cloud control. Though the cloud is distributed to different physical locations, customers benefit from a single point of control. This prevents any discrepancies or conflicts in cloud management.
- Flexibility. Distributed cloud, deployable virtually anywhere, guarantees location flexibility.
- Compliance. Since the entire stack is distributable anywhere, customers more easily meet compliance demands dictating that certain data must reside either on-premises or in a specific location.
- Network reliability. The risks of a global outage, while not entirely mitigated, are decreased. Each distributed cloud operates individually on local subnets, preventing an entire network from failing when one system fails.
Drawbacks and challenges of distributed cloud
However, distributed cloud, in large part because of its complex nature, presents several challenges:
- Cost escalation. Setting up an entire cloud system every time a customer deploys to a new location is costly. Overheads also increase as maintenance costs multiply for each new system.
- Data discrepancy. A distributed cloud infrastructure faces data integrity issues. Ensuring all nodes always have access to data while also making sure everyone has access to the most recent version of that data is difficult. Replication of data across nodes, if not executed swiftly, leads to data discrepancies.
- System inconsistencies. Designing and maintaining a system comprising multiple independent sites is challenging. Ensuring consistency across nodes, especially when troubleshooting or making changes to the system, adds to the difficulties.
- Security misalignment. The ability to deploy almost anywhere has undeniable benefits. However, distributed clouds sometimes reside within an infrastructure subject to security requirements that don't align with those in other locations.
- Data movement. Though distributed clouds easily move data from one location to another, that's not always the desired outcome. Backup or contingency plans must ensure any geographically sensitive data stays where it belongs.
Main examples and use cases for distributed cloud
In effect, distributed cloud combines the concepts of public cloud and edge computing, improving computing services by bringing them physically closer to a variety of clients, including those seeking the following:
- Content delivery. Content delivery networks benefit from distributed cloud by storing and streaming data from locations closer to end users. Data-intensive services -- video streaming in particular -- benefit from decreased latency, as do gaming services and internet of things applications.
- Regulatory compliance. Distributed cloud is available in almost any physical location necessary to meet any compliance-related mandates.
- Continuity and recovery assurance. With proper precautions in place to mitigate data integrity issues, distributed cloud, thanks to its inherent redundancies, supports business continuity as well as disaster recovery.
- Faster data analysis. Arguably today's hottest tech-sphere topic, artificial intelligence and machine learning applications require distributed cloud's time-sensitive processing and analysis of data to deliver lower latency.
- Streamlined workflow. In any scenario involving large amounts of internet traffic, distributed cloud improves outcomes by dispersing workloads across multiple nodes. For example, a digital storefront is slammed with traffic during an online sale, but distributed cloud ensures customers aren't delayed at any point while shopping the site.
Traditional cloud vs. distributed cloud vs. hybrid cloud
While distributed cloud incorporates elements of traditional and public cloud and hybrid cloud shares similar principles, there are key differences.
Traditional cloud infrastructure is limited to a single physical location. Distributed cloud is deployed to multiple locations even on existing infrastructure, decreasing the distance between cloud and client, lowering latency, and providing improved performance over traditional cloud.
Hybrid cloud systems also consist of multiple clouds, but hybrid cloud combines both public and private clouds, such as on-premises data centers. They also can't offer distributed cloud's centralized control system, which is a simpler management solution.
What to look for in a distributed cloud provider
Ultimately, the criteria for picking a quality distributed cloud provider match those used when selecting any cloud service provider, traditional or otherwise. Cost, technology choices, security, governance policies and reliability remain significant factors.
Distributed cloud is set up wherever it's needed, so data center location is unimportant. But customer support is important.
When selecting a provider, such as cloud offerings from IBM, VMware or Akamai, recognize that the complexities and wide-reaching nature of a distributed cloud network require a swift yet steady hand in coordination, logistics and ongoing management of the infrastructure. Seek a proven, reliable provider with the necessary resources to resolve issues as they arise and be aware they might pop up anywhere or everywhere the distributed cloud resides.
