Optimizing Docker Builds and Debugging AWS Lambda for Glowing Telegram | Ep. 195
3In this continuation of our Glowing Telegram series, I focus on refining the backend infrastructure for the project by optimizing Docker builds and debugging AWS Lambda functions. This is Part 2 of the livestream, and Episode 195 in our full VOD series on YouTube.
I begin by analyzing ways to improve the efficiency of Docker builds, specifically by targeting individual binaries using multi-stage Docker files. This approach reduces build times for individual services and lays the groundwork for consolidating multiple Docker files while maintaining configurability for each service.
Next, I dive into fixing issues with AWS Lambda functions handling video metadata and JSON playlist generation. Through detailed debugging, I identify and resolve several roadblocks, including missing transcoded data, DynamoDB query issues, and improper cloud caching behaviors in CloudFront that affected playlist functionality.
In addition to solving immediate bugs, I discuss strategies for designing reliable backend architecture with AWS services. Topics include managing API rate limits, orchestrating video uploads to YouTube using AWS Batch and Fargate, and designing an efficient, scalable queue system leveraging DynamoDB. These discussions set the stage for automating YouTube uploads while ensuring persistence, scalability, and adherence to quotas.
Join me as I plan the next steps in this project, including building out an infrastructure to automate YouTube uploads. This video serves as both a troubleshooting session and a roadmap for upcoming features in the Glowing Telegram project.
For more streams where I tackle backend development challenges and coding workflows, check out the links below:
🔗 Check out my Twitch channel for more streams: https://www.twitch.tv/saebyn
GitHub: https://github.com/saebyn
Discord: https://discord.gg/N7xfy7PyHs