Nanoservices live within the sizing constraints of Microservices. Smaller size of modules enables the service to maintain and change. Therefore, a system composed of Nanoservices can be easily extended.

In a typical domain driven system, each class and function can be modelled as a separate Nanoservice. This leads to an increase in infrastructure costs such as that of application servers and monitoring solutions. Since implementing a complete business solution using Nanoservices might involve deploying several hundreds to a couple of thousands of Nanoservices, the infrastructure costs on the desired cloud platform need to be as low as possible. In addition, Nanoservices should not be long running or resource intensive.

Nanoservices require a lot of communication among themselves. This may lead to degraded performance. On certain platforms, Nanoservices may share a process, which may cause resource starvation and take away technological freedom. This approach however, can reduce the overhead of inter-service communication.

Before you plan to commit to a platform for hosting Nanoservices, the following objectives should be considered:

• Cost of infrastructure for hosting Nanoservices should be low. The platform should support rapid deployment and monitoring of Nanoservices.
• The Nanoservices platform should support dense deployment of Nanoservices.
• Inter-service communication will degrade performance. A highly reliable and efficient means of communication would ensure that the performance remains within acceptable limits.
• Nanoservices should be independently deployable and the platform should ensure that failures are not propagated between Nanoservices.
• Nanoservices are limited in terms of choice of programming language, platforms, and frameworks. You should study technical feasibility of building Nanoservices on a platform before committing to it.