Newbie Scaling problem/question
I'm a golang/devops newbie.
I have a Node JS web-application and chatbot that uses this repository as the golang image server to handle all the images processing (with file cache system).
The image server is deployed in 2 instances and load balanced.
I predict that my current setup serves/handles about ......... 500,000 to 1,000,000 images/requests per day in various sizes ... Hard to tell what the number is exactly.
About 3 weeks ago, I noticed that a lot of my images were loading extremely slow and turning up broken. At that point I upgraded both my instances from micro to 1 vCPU, 3.75 GB mem. That actually solved the problem easily.
Fast forward to now, I'm starting to notice some slow loading images again and saw the broken images issue once or twice only.
Could I please get some advice and tips from you for dev-ops and scaling?
I have never really had to handle scaling before and I'm not sure how and what data to look at to measure if my instances' resources are sufficient.
For example, this may sound really dumb but when I look at the monitoring section in Google Cloud Platform, the CPU only peaks at 3% to 4% max, and I'm unsure if I'm looking at the right numbers.
Thank you kindly in advance.