Subscribe Us


 

Kubernetes for Edge Computing Paving way for empowerment in the next decade!!!

Kubernetes for Edge Computing

Paving way for empowerment in the next decade!!!

 

Pat MacDonald couldn't have been righter whilst penning down the acclaimed lines, "The future's so bright, I gotta wear shades…". 36 years thereafter, we are yet experiencing the same, firsthand! And it doesn't just stop here! Things are going to go way better and much higher from here.

 

 

Empowerment till now: Till just a few decades back, talks of the gap between the haves and haves-not used to carry some substance. Power and opportunities used to be looked upon by the common masses as more of an outcome of luck than effort. Enter the humble Silicon, and today, the playing fields have been levelled like never before! With the empowerment of having access to any sort of information in an organized manner through the medium of text, images, videos and what not, the above-mentioned gap has merely begun its journey due south! Whether it be about gaining popularity, or sharing one's ideas/flairs with the world, or be it automating any business or personal process, or even be it the comfort of having stuffs delivered at one's doorstep, computing technology has never prejudiced while empowering a person.

 

 

Empowerment henceforth: The millennium has turned 22! And with IOT and Data Analysis becoming an integral part of our lives, we can but merely wonder how brighter would the future shine! And the forebearer in this is what is termed as "Edge Computing".

 

 

Cloud computing: Imagine your online searches. Every search that you do becomes a data point for the search engine and its associates. Not just the searches vaguely, but the exact key-words that you entered, the amount of time you spent on a particular webpage, the links that you clicked, so on and so forth. Each and every one of these details are not merely stored in the servers in the form of data but are continually analyzed to create a virtual doppelganger of you. Surprised? Well, you would not be if you recollect how every social media starts showing you ads on shoes right after you had searched for 'shoe stores near me' or how the prices of flights between Narnia and Gotham keep on increasing in a matter of a few minutes when you search using same or similar keywords many times. Each and every place that you have ever visited (after having a google account active on your phone, that is) is stored in the Google cloud. The point here is that your data is being stored and analyzed at a remote place, called as the cloud. Now this cloud system has been a pretty wonderful tool for data


storage and management over the traditional tools (compare your computer file system to Google Drive or OneDrive storages). Yet, there are certain necessities that prompted mankind to look beyond the cloud system as well. What might be these necessities you ask?

 

 

Edge computing:

Challenges in cloud computing: Imagine you using an IOT home security system over your mobile phone while you're away. In order for the data to be stored in the cloud, you would require a great amount of bandwidth. The Almighty forbid, but suppose there's a robbery taking place at your place! Using a remote server to store, sense and then act on the data would lead to criminal wastage of time. This delay is termed as latency. Moreover, during such an  occurrence, if  there is  any break down in the server system due to equipment failure or, maybe, power outage, the resilience of the server has to be high enough to quickly recover from the setback and safeguard your dear home. Finally, there can be data sovereignty concerns in that part of the planet where you reside. This has to do with protecting your interests while capturing and making use of your data.

Overcoming challenges through edge computing: How do we take care of these shortcomings you ask? Well, if instead of having to rely on the cloud system  for storing the data and then acting upon it, if the data can be stored within your place while granting it the power of attorney to take immediate remedial action as and when needed, would that not be wonderful? Sure would be. And folks, this is what edge computing is all about. Simply put, it is the phenomenon of bringing computing services such as data analysis, closer to you, the consumer, or, to the source of data, your home being the one for the illustration in question. Though, edge computing, by no means, ought to be seen as a replacement for cloud computing. It is more of an extension of cloud computing. Since the data is being stored and analyzed near the data source, the latency reduces drastically. For the same reason, resilience, bandwidth and data sovereignty issues also cease to bother. Some examples of edge computing:

·       Remote monitoring in the oil and gas industry

·       In-hospital patient monitoring

·       Traffic management

·       Smart homes

 

 

Requirements for deploying edge computing: Based on the discussion above, it becomes quite clear that in order to deploy edge computing for personal or business purposes, hardware requirements are quite heavy. Some of the primary requirements for such a system are:


·       High performance

·       Versatility in mounting

·       Rugged construction

·       High storage capacity

·       Wide power range

·       Wide input-output connectivity options

·       Hi-speed internet

Since the explanation of these requirements lie outside the scope of the discussion in question, we shall discuss these in detail when we discuss the set-up for edge computing in a separate write-up.

Apart from the ones stated above, there are few requirements that an edge computing deployment has to serve:

·       Scalability – Imagine you have an edge computing system deployed for your café. Eventually, as you grow and franchise out your café, the locations are going to change and the functions that your edge computing is currently handling would have to be handled multi-fold now. This would require tools that would be able to automate and orchestrate the managing of the processes while keeping things architecturally consistent.

·       Portability – Since the edge computing system has to be near the data source, it has to be portable.

 

 

Layers required in edge computing: An edge computing is required to have at least two layers. The purpose of the same is to simply facilitate smoother operations of the organization and efficient working of the developer. The layers are:

·       First layer: The first layer takes care of the computing, storage and bandwidth requirements.

·       Second layer: In the device layer, the workloads in the data center and the public clouds are orchestrated. The tool that has become most popular for this purpose is the 'Kubernetes'.

 

 

Kubernetes: The word 'Kubernetes' comes from a Greek word which means a boat steersman. The short form of this term is K8s (the figure 8 representing the number of letters between the first and the last alphabet, i.e., between K and s). The computing resources in edge computing are stored in the form of containers. The primary task which a K8s performs is to manage these containers. The tasks that a K8s performs are:

·       Auto-healing: At times, containers may fail due to certain reasons. It becomes imperative that these containers are restarted smoothly. Again, at times,


containers do not respond and it becomes necessary to kill the container. These tasks are performed by the K8s.

·       Auto bin packing: Nodes are virtual and/or physical environments  in  which the K8s specify the amount of storage and processing power that has to be allocated to a container.

·       Selection of storage: There are many cloud storage systems available. The K8s provides with the flexibility of being able to choose the one/s most suitable for you.

·       Optimizing docker and container usage: A docker image is a set of instructions given to manifest a docker container. It is through the K8s that the minimum number of such dockers and/or containers are used while not over burdening them with more functions that they can bear.

 

 

Final Words: With the rising usage in the 5G technology, the demand and scope of Kubernetes for edge computing is slated to grow. Kubernetes are already  being used in around 60% of the edge computing tasks and would capture more share subsequently. The technology of edge computing is slated to open new doors for our personal and entrepreneurial lives. We simply have to watch it unfurl!

 

 

 

Penned by:

Dhruv Dev Dubey

 ddcontentwriter@yahoo.com

Post a Comment

0 Comments