It seems like every other day there's another post on some social media platform from someone trying to grok a massive bill from any one of many observability platform providers. I've never been convinced of the business case for using any of these services and in this talk I'll go through some of the considerations I take in to account when deciding what to deploy and we'll crunch some numbers to see why it often makes sense to do it yourself, instead.
Some topics we’ll cover.
Cost - Transparancy Cost - Predictability Metric Quality Speed Reliability Feature Sets We’lll discuss the role of centralized logging and it’s intersection with application metrics
The take-way’s will be a broader view on how to approach setting up or migrating your observability stack and an example stack that I have have experience running along with some real numbers from it.