Moving services to the cloud is a growing trend. For new companies, it's not unusual to have practically all their IT with one or several cloud providers. Will 2019 be the year of the big jump to the cloud? This is what IDG “Cloud Computing Survey” seems to indicate. It maintains that 73% of companies already have one or many services in the cloud, and up to 38% are thinking of migrating their services.
Several factors must be kept in mind when moving services to the cloud. First, legal requirements must be verified and the economic return proven. One thing many companies have done is to map out which, and in what order to migrate their services. It’s not the same to migrate a monolithic app designed 15 years ago as it is to migrate a recent one. Therefore, each one has to be treated separately.
For traditional services, there are a proliferation of tools that make these migrations easier. Velostrata, a company recently bought by Google, is a good example. It enables moving virtual machines from a VMware environment to the cloud in a matter of minutes. Moving virtual machines to IaaS, is probably not the best way to optimise resources, but its similarity to traditional IT justifies the use of this approach.
At the same time, one of the biggest advantages of Google Cloud and other platforms is Platform as a Service (PaaS), where there is no need to install and maintain databases, for example.
One of Google Cloud's points of differentiation is data management. This is facilitated through its services, the deployment of Data Warehouse, NoSQL or transactional databases, characterised by elasticity, performance, availability, and price leadership in the market.
If the alternative is to create a new application instead of migrating an existing one, cloud providers facilitate the use of new technologies like container orchestrators (Kubernetes). For these cases, Google Cloud is betting on open software, with the goal that customers can avoid being locked into one specific provider, (vendor lock-in).
In this area, Kubernetes, the most popular container orchestration community, was launched some years ago. In 2018, two new initiatives were created. The first is called Knative, and it's a development and execution environment to build scalable and simple applications with the possibility of delivering them to any cluster of Kubernetes. The second initiative is called Kubeflow and helps manage Machine Learning models in their creation, deployment and maintenance.
When talking about Machine Learning, it’s essential to point out that it’s one of the biggest growth areas in new technologies. Tools like Tensorflow and Auto-ML simplify the design and implementation of solutions.
Finally, let’s highlight voice as a new interface, an area that in 2019 is continuing to grow. Thanks to the maturity of the tools and services of Machine Learning, creating conversational robots are within reach of everyone.
In conclusion, if you're looking to advance in the IT world, my recommendation is to get training in technologies like orchestrated containers and Kubernetes, or enter the fabulous world of Machine Learning through certifications offered by Google: