Image Processing Service: From On-Premise To Cloud Function With Golang

on YouTube ↗09m

" summary

Image processing moves from a monolithic Python app to small Go functions behind an API Gateway, writing standardized files to Google Cloud Storage and serving through CloudFlare CDN; system handles 200 million requests total, ~360k new images/day and ~90 GB/day growth. Uploads respond 99% under one second and reads are typically under ~600 ms. Problems solved: single-server storage limits and outages; outcome: auto-scaling Cloud Functions, fast cached reads, and simpler image resizing pipeline.

AI-written summary · about

" chapters (8)

1. 0:00Start
2. 1:08OVERVIEW
3. 3:30CHALLENGE
4. 4:51ST VERSION - PROBLEMS
5. 5:49RD VERSION - 2018
6. 6:18RD VERSION - PROBLEMS
7. 6:46TH VERSION - 2019
8. 8:22THE FUTURE

" related talks

• 02h 47mBuilding a Photo Sharing Service with Google Cloud Serverless TechG. Laforge · Devoxx France 2021
• 43mGoogle Cloud's new serverless featuresGuillaume Laforge · DevFest Nantes 2018
• 46mHighly Scalable Image Storage with AWS ServerlessVadym Kazulkin, Firdaws Aboulaye · GOTO Amsterdam 2024
• 40mWhat's new in Google Cloud serverlessGuillaume Laforge · DevFest du bout du monde 2019
• 49mGet silly with serverlessBret McGowen, Guillaume Laforge · Devoxx Belgium 2017
• 50mDeveloping Highly Scalable Image Storage Solution with AWS ServerlessVadym Kazulkin · GOTO EDA Day London 2024
• 35mIt’s Easy to Say ServerlessAlessio Coser, Daniel De Paoli · Incontro DevOps 2017
• 43mWhat's new in Google Cloud ServerlessGuillaume Laforge · DevFest Paris 2019