Happy AWS Summit Day! It’s been a long day watching several events online, jumping through several different topics. So I wanted to give a quick stop tour of the events that I was able to watch live today!
Web and Mobile Application Development Powered by AWS Amplify
Sini Tinfors (Startup Solutions Architect)
By using AWS Amplify most of the work is done for you. The developer can use the console to create the structure for the data model. Once this is set up and deployed, behind the scenes Amplify will create an AppSync endpoint to the GraphQL API.
While waiting for that to be deployed, the developer can continue working on the front end. The example in this demo was created with create-react-app.
Once Amplify has deployed the data model, it will provide the user with the console instructions to clone it. And also code snippets which can be added into the front end code to link the two repositories. There are code snippets for each CRUD instruction which may be needed.
This one was a great one to start the day as the demo really helped show how easy Amplify is to use. I have so many ideas which I want to try out, and getting more insight into Amplify has encouraged me to give it a try.
Selecting and Optimising Amazon EC2 Instances
Alpana Goolab (Solutions Architect, AWS) & Jonathan Harvey (Head of Common Platform, ITV)
This talk was based on how ITV have optimised their usage of Amazon EC2 Instances. With the need for scalability, the session highlighted the different costing benefits of different types of instances AWS has available:
- On Demand Instances – pay for the compute capacity by the second, which is ideal for spikey workloads and does not require long-term commitments.
- Reserved Instances – 1 or 3 year commitments which is good for committed and steady states of workload. It can provide discounts of up to 72% compared to On-Demand instancing pricing.
- Savings Plans – this is the most flexible option in terms of pricing, based on a commitment of 1 or 3 years.
- Spot Instances – the cost of these instances are up to 90% less than Reduced Instances.
Scale with Containers in the Cloud
Brajendra Singh (Sr. Solution Architect, AWS) & Alex Howard Whitaker (Principle Cloud Services Engineer, Ocado)
The increase in online shopping due to Coronavirus and lockdown meant that Ocado had to ensure the company did not suffer any downtime from the increased demand. So the scaling capabilities which comes with AWS Containers was a huge advantage. However, the interesting part about the discussion was how Ocado set up the infrastructure so that the scaling etc could happen through an abstracted layer, so developers could continue to work as normal without interacting with the DevOps side.
Another factor was how Ocado, and many other companies, rely on CI (Continuous Integration)/CD (Continuous Deployment or Delivery). Due to the high number of deployments per day, there was the requirement to keep track of the “health” of the different instances. In AWS this is done by centralised logs through ElasticSearch and Kibana, along with CloudWatch metrics.
Improving Code Quality and Application Availability Using AWS AI Services
Ozioma Uzoegwu (Solutions Architect, AWS)
This was was particularly interesting as I had never looked into A Cloud Guru before. It was launched in 2019, and it uses AI to provide feedback to pull requests. These pull requests can be submitted through several platforms including GitHub and BitBucket. The advantage of using AI over other methods of automated methods is that it can keep up with the latest trends by learning from feedback on AWS open source projects. While other methods will typically provide quite high level feedback.