Five Challenges with Cloud of Things Analytics

COT WP
Cloud of Things analytics, which marries the capacity of the public cloud with the reach of IoT, has huge potential for operations teams. Anomaly detection can ensure service quality, trend analysis can improve customer service, predictive analytics can avoid costly downtime and prescriptive analytics can automate intelligent real-time action. But transitioning from a datacenter-constrained, on-premises analytics system to an elastic cloud-based platform can be daunting. Let’s take a look at the main challenges implementing Cloud of Things analytics.

Provisioning
Configuration and installation of cloud-based analytics can be complex. Unfortunately, an on-premises application cannot simply be moved to the cloud. Proper connectivity for data ingestion (volume and velocity) must be established, and perhaps more importantly, firewalls and other access rules must be re-established in this new environment. It’s also unlikely that the compute and storage framework will be an exact match, so certain configuration parameters may need to be tweaked as well. For example, batch sizes may need to change based on connectivity and compute resource changes, or data persistence rules may need to be adapted.

Application delivery
The analytics application needs adequate delivery, maintenance and troubleshooting. This ensures a robust experience for users without unexpected round-trip delays. Unlike a private datacenter, organizations cannot install traditional load-balancer appliances which might handle this function. There are software-based and virtualized ADCs, and public cloud providers have their own rudimentary offerings, but application delivery and management will need a fresh approach in the cloud.

Data security
Closely linked to delivery is security. There are two elements here, data security and access, which we’ll cover next. As recently as 24 months ago, security was a major concern for organizations transitioning valuable data in the cloud. What if it gets hacked? That’s still a significant concern for some sectors, but many enterprises have realized that Amazon, Google and Microsoft have deeper security resources than they have in-house. So their data might actually be safer in the cloud than in their own datacenter. That said, servers still need to be hardened, penetration tests done and additional security measures (such as web application firewalls) installed. The cloud-based analytics system needs to be locked down just like other cloud apps.

Access
The second facet of security is access—who can log into the system. User authentication, granular permissions and password policies are all standard access control disciplines, but they become more important when an application sits beyond the corporate firewall. Insecure passwords and unchanged root admin accounts can expose analytics systems and all the company’s valuable data to unauthorized access. Consideration must also be given to mobile device management. If the data is valuable and needs to be acted upon in real-time, then mobile access is desirable. However, it comes with additional access risks given mobile devices are easily lost or compromised without a mobile management platform. For an on-premises system, which wasn’t reacting in real-time, this might not have been an issue or even an option. Not so with Cloud of Things analytics. Mobile access is the default.

Regulations
Public cloud providers offer application and data hosting in multiple geographic regions. Data hosted in a country such as Germany will come under German law, which can differ greatly to American legal requirements. Consideration then must be given to the regulatory environment when it comes to Cloud of Things analytics. HiPAA, PII, GDPR and other data regulation measures can affect every company, regardless of size or sophistication when hosting data in the cloud and serving it to users worldwide. The organization might be outsourcing data storage and compute, but not its compliance responsibilities.

While it may seem daunting, the potential of Cloud of Things analytics is significant, and the journey worthwhile for forward-thinking organizations. In the next in our series, we’ll look at the advice we give to organizations implementing a more robust analytics framework.

Would you like to learn more about Cloud of Things? Download Our White Paper!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>