Logs are very important. If you're running a live app you'll need them to find bugs. It's especially important when you're running machine learning in production because the new data that comes in can indicate if the performance is changing over time. You may even need to re-label new datapoints that come in.
Logs also serve different purposes, which is why you typically see multiple places where logs need to be sent to. Some logs need to go to could storage, while others need to appear in reporting tools like the elastic stack. Setting up logging appropriately can get involved, especially when you're running multiple apps.
If this challenge sounds familiar, you may be interested in vector. It describes itself as "a lightweight, ultra-fast tool for building observability pipelines". We found it to be a like-able tool, which is why in this series we're going to show you how you might be able to use it.
If you want to tag along, you will need to install it first. For Mac the installation is done via homebrew.
brew install vector
For other platforms you'll want to check the installation docs.