At 8.30am, on Saturday morning (August 19), the first day of AWS DevDay 2017 officially took place in Hanoi. This is the technology event organized by two leading technology groups: Amazon Web Services (AWS) and FPT, particularly for developers who are interested in Amazon cloud computing, and how to maximize resources on Amazon Web Services.
Developers, speakers, and reporters are at The Vuon – Luxury Garden Office to check in and be ready for the conference.
Accordingly, each successful application received a QR code to “through the gate”, each participant will receive $50 credits to AWS account in order to practice using AWS tools in the afternoon.
Speakers from AWS and FPT are early present to prepare for the presentation and demo at the event.
Developers who are interested in cloud computing settled in to prepare for the first session.
The event started with Mr. Parijat Mishra – A manager, Solutions Architecture at Amazon Web Services. He tells how to build an application architecture on AWS.
At the beginning of the presentation is the story of Amazon’s growth in reality. Amazon begun as a Monolithic application which is based around data, multiple functions.
After some time, some problems were observed: the front-end evolved into a massive, complex application with lots of databases behind it, and the data store was very hard to change. This slowed down the rate at which Amazon could evolve.
In order to solve this, Amazon has carried out a “transformation”, moving from Monolith to Services. Each “service” had its own team – of engineers and business people – that owned the development, deployment and roadmap for the service. The results showed that the code quality increasing significantly.
The Amazon before and after the new model is illustrated by the figure, which can be seen in the right-hand model showing more flexibility than left one.
However, with services model, Amazon also encountered the same problems such as having to scale all applications equally while there were applications that required more scale than others. And true Micro services is the solution Amazon chose to solve this problem. True Micro services are based on function, did not share data, and scaled independently. It means that when a service is in the process of scaling, the other services would still be able to function normally, without being affected.
Participants have 15 minutes for networking, tea break and relaxing.
At the second session, Mr. Seon Yong Park – Developer Specialist SA at AWS delves into the features of AWS SDK (Software Development Kit on AWS).
The AWS SDK supports many programming languages such as Java, .NET, PHP, Python, Ruby. That is why Amazon is always proud of telling about their flexibility. In addition, Mr. Seon Yong Park also emphasizes the importance of credential when starting programming to ensure security for application.
The third session continues with the knowledge of getting started with Serverless and AWS X-Ray shared by Mr. Parijat Mishra. At his presentation, he explains the components of Serverless, and why programmers should care about it during programming.
Specifically, he goes deeper into the Lambda function, the compute option of the serverless offering at AWS with object storage using S3, NoSQL DB and real time streaming using Kinesis. It can be applied in many languages, and tagging for Lambda helps take advantage of administration benefits such as tracking, or separate costs for different groups and projects. From there it is possible to compare costs and benefits to make “investment” decision in the future.
The problem with Lambda lies in how to ensure that when changing a function without affecting the rest and how to combine functions with each other, or which function to use when using too many functions.
In response to this question, Amazon has created a sample architecture based on the best practices. Programmers can refer to sample applications on the AWS site.
Demo on Lambda function, the speaker goes into the analysis of the application managing photo album. Lambda function is used to detect objects in images after they are loaded on S3, based on the object’s particular properties to find where the object is.
Attendees and speakers have buffet lunch at the Eastern Garden Room, The Vuon.
After the lunch, developers return to the hall to continue the workshop and hands-on lab. The afternoon session starts with the presentation of the AWS Solution Architect, Ms. Rohini Gaonkar, with DevOps on AWS.
DevOps is a combination of developer and operator. Traditionally, developers and operators operate separately, the developers who focus on coding usually prioritizes the speed, and the operators is more concerned with the stability of the application. DevOps is a combination of these two, through the Continuos Integration/Continuous Delivery (CI/CD).
According to Ms. Rohini Gaonkar, DevOps is a combination of tools, technologies, and development processes. Devops can refer to two major areas: Application Development (including: code building, unit testing, packaging, deployment), and Infrastructure Management (including: provisioning, configuration, orchestration, deployment).
DevOps model relies on effective tooling, high velocity. AWS provides services that are designed for DevOps and that are built first for use with the AWS cloud. These services help use the DevOps practices described before.
AWS CodeStar is a unified user interface to quickly develop, build, and deploy applications on AWS. It helps rapidly orchestrate an end-to-end software release workflow using these services. Using CodeStar, owners can add and remove other team members, contribute code to the project repository, grant or deny other team members remote access to any Amazon EC2 instances running Amazon Linux associated with the project, configure the project dashboard, and delete the project.
And contributors can add and remove dashboard resources such as a JIRA tile, contribute code to the project repository, and interact fully with the dashboard. Cannot add or remove team members, grant or deny remote access to resources, or delete the project.
Mr. Seon Yong Park defines Artificial Intelligence – AI through Turing test: There are machine and human over the wall, evaluator asking several questions. If evaluator cannot tell machine from human, the machine is said to have passed the test. That’s one way to define AI, he said: “Knowledge changes day by day. AI is not only about Intelligence but also Interaction.”
Machine learning and deep learning are what whole the world is talking about, and Amazon is not out of the hot topic, when offering three services including: Amazon Recognition (a deep learning image analysis service), Amazon Polly (a Text-to-Speech (TTS) cloud service that converts text into lifelike speech) and Lex (a service for building conversational interfaces using voice and text).
He also gives some examples of Recognition by giving a photo and then extracting information about objects appearing in it. Besides, he also describes how to train a neural network in deep learning.
Mr. Ho Minh Hiep – Solution Architect from FPT.AI, introduces FPT Speech Synthesis, text-to-speech product using deep learning technology, running on AWS platform. It is said to be the best today with many features, voice, continuous improvement and especially API is public to the community.
According to Mr. Hiep, AWS provides many tools to achieve business requirements with optimized cost. “EC2 with auto-scaling is still a very solid tool for peaky traffic. Lambda is a powerful tool to scale up application seamlessly. The cost should be measured carefully, however.” said he.
See more about FPT Speech Synthesis HERE
Time for tea break and for speakers and attendees to take a rest.
The practice takes place in more than an hour with the concentration but excitement. Experts from AWS and FPT directly instruct programmers how to use tool sets to optimize resources on AWS.
In August 20, AWS DevDay 2017 continues at Meeting Room B02.09 – HUTECH (HCMC University of Technology), 475A Dien Bien Phu, Ward 25, Binh Thanh District, Ho Chi Minh City.