Why IoT Needs a Cloud-Native IoT Messaging Service

uriotnews

Today, there are two main challenges to building an IoT application that provides bi-directional communication between IoT devices and IoT cloud platforms:

  1. The big cloud providers use IoT solutions that lock-in customer devices to the individual cloud provider.
  2. Managing and deploying a messaging service requires considerable expertise and resources. These factors are serious barriers for many companies that want to deploy production IoT applications.

You may also like: Multi-Cloud for IoT

Big Cloud Lock-in With MQTT

Amazon, Microsoft, and Google all have proprietary IoT platforms. A key feature of these IoT platforms is managing data messaging between IoT devices and the cloud. The good news is that all of the big cloud vendors support MQTT as one of their messaging protocols. The bad news is that none of the big cloud vendors support the complete MQTT specification and none have announced plans to support MQTT 5 specification.

Most of the big cloud vendors implement a modified MQTT version that locks devices into only being able to communicate with the big cloud vendor MQTT service. These big cloud vendors talk about MQTT but don’t deliver the value of an open standard.

Vendor lock-in is always a risk for customers. For IoT applications, the risk of vendor lock-in is even greater. Typically, IoT devices are deployed outside of the enterprise, have long product life cycles, and involve a significant number of deployed devices. If your IoT devices can only talk to one big cloud IoT service, you are tightly bound to whatever that IoT service decides to offer.

We believe IoT devices need to be cloud-agnostic. It should be possible to direct IoT device messaging data to any cloud vendor or even multiple cloud vendors. We live in a multi-cloud world, so it is important that IoT applications can exist in a multi-cloud world.

Reduce the Complexity of Production IoT

Our industry needs to work at reducing the complexity of deploying and managing production IoT systems. IoT technology is far too often geared for doing quick proof of concepts; however, once a PoC is moved to production the complexity of managing the system can be overwhelming.

Few organizations have the people and processes to manage the scalability and availability business-critical IoT applications require. IoT applications must be able to scale up and scale down seamlessly to accommodate spikes in usage. Business-critical IoT applications often have service level agreements (SLA) that require high availability of cloud services.

Delivering the desired level of scalability and reliability can often be a limitation for deploying an IoT application into production. The costs to implement scalability and availability for an IoT application can be considerable.

Options for MQTT Cloud Deployment

Currently, customers that want to deploy an MQTT-based system into production have two options: 1) Use a big cloud service that does not completely support MQTT and accept vendor lock-i2) Manage an MQTT broker cluster themselves. We believe the industry needs a third option: an MQTT cloud service that is 100% MQTT compliant and doesn’t require a vendor SDK to connect a device. We call this third option a ‘Cloud-Native IoT Messaging Service’.

Cloud-Native IoT Messaging Service

Our definition of a Cloud-Native IoT Messaging Service is as follows:

  1. The standards-based offering doesn’t rely upon the services of a particular cloud vendor.
  2. Transparent, elastic scalability that can accommodate peaks and valleys in telemetry data traffic from IoT devices.
  3. Transparent fault tolerance and high availability that fulfills the service level agreements business-critical IoT applications require.
  4. Ability to run on different cloud platforms.
  5. Open API allows for the integration of the data with other cloud and third-party services.

Open Standard Implementations Required for IoT

For the IoT industry to reach its full potential, customers need to have the flexibility of vendor-neutral standards and 100% compliant implementations. The current cloud implementations of MQTT are leading to vendor lock-in that will limit the future growth of IoT applications and future potential for the industry. 

HiveMQ has recently launched HiveMQ Cloud, a fully managed MQTT platform that provides a cloud-native IoT messaging service. HiveMQ Cloud is based on a 100% compliant implementation of MQTT. We hope and expect other cloud implementations will become available to further the success of MQTT as the defacto IoT messaging protocol.

Further Reading

Cloud-Native Databases and Why You Should Use Them

How Cloud-Native and Kubernetes Make Application Development Easier

Securing Cloud-Native Apps

This UrIoTNews article is syndicated fromDzone