Posted by John Crawley | 11 February 2026 at 2:06pm
Tags: Ansible AWS AWShorts Bash EC2 episode 2 episode 3 Glue Terraform | Categories: AWShorts SeriesAs mentioned in my previous blog post, I do want to continue this YouTube series onwards, but couldn’t in the previous 9 months due to events and personal reasons in 2025, which I feel is a shame because it was fun just making small fun projects, demostrating them on YouTube then upload the project to the GitHub repository.
I want to talk about the next two episodes I’ll be possibly making. Normally I rather upload the episode then talk about it but just too excited so this information is subject to change since I’m researching these two episodes currently & planning it out:
Episode 2 will continue with IaC using Terraform to initialise EC2 instances (& destroy when ep’ done), then using Bash to automate getting the IP addresses for Ansible inventory which I will use to configure the EC2 instances and application deployment.
I want to break away from making the series feel too much like a DevOps series (would be great but not the point, the point of the series is just to use AWS to make projects I think would be interesting to do with various AWS services). In this episode I plan to use a huge dataset with serverless data pipeline by uploading the dataset to S3, once uploaded it will automatically trigger AWS Glue to detect the schema and prepare the data where it will finally go to AWS Athena to show insights into the data where I can instantly pull insights using SQL and no servers, just using AWS services together to turn raw data into real insights in minutes.
As also stated in my previous blog post. While this is still true to an extend now, I might be able to record the episodes on Saturdays but in order to continue this series with what I currently have in terms of a quiet place, I will need to change the format of the video, so in episode one I recorded my voice while doing the episode, I will now need to record the footage first and record my voice in post-production (I prefer the format of episode one, less video editing and planning). This means the short window I have for a quiet place to record, I can utilise the time to record the commentary. Also, I may have lost my microphone so audio quality may go down (I apologise) until I can get a new one at a later point, unsure when. Make do with what I can.
For episode 2, I’m finalising the research and planning (the idea kept changing previous week). I’m unsure when I’ll record the footage, I’m aiming to be very soon (could be next week but don’t hold me to it) and then on the Saturday following that, I will record my voice over the video and should have the episode uploaded on Sunday or Monday night.
I could be wrong but for episode 3, it should be a little faster to produce since it’s creating an automation pipline, upload dataset then see it in Athena but if possible, I can spend a day in Starbucks recording both episodes. I can’t see why both episodes shouldn’t be released before the end of this month but let’s see.