Not every problem is a nail. Similarly, serverless isn't right for every situation. This particular reason stood out to me.

...you have a tight timeline to deliver a product and a lot of expertise in a different technology.

This is something I've observed personally. I meet with many startups and while there's a lot of interest in serverless in the startup world, many I talk to choose containers because they know the technology and are comfortable with it. They see serverless as undifferentiated heavy learning.

You can disagree, but the ones I've spoken to felt their time was better spent on product development than learning serverless. This has posed an interesting challenge and fueled repositioning on our end around our services.

You may be like me and have felt in the microservice era that CPU profiling wasn't the most worthwhile. If you could quickly scale your service vertically or horizontally out of your problem it was possibly more efficient work to make a scaling change than to write more efficient code.

But with serverless there's an argument to be made that CPU profiling becomes important again. See how IOpipe helps you write more efficient code with its profiling feature.

Been looking a bit at learning Go recently. As an operations person (it's in the company name!) I think it's important to be multilingual to be as effective as possible. I'm very comfortable with Python. I've tried learning Javascript but I've found it difficult to understand. Go has slowly become more interesting to me.