But I'm not quite ready to stop yet, either.
And why do we feel like such a discontinuous break with the decades of prior art in monitoring, logs and APM?
friends, man 🐝💜
First, simple tool illiteracy. I've been an ops eng since I was 17, but monitoring always bored me.
It was an almost religious experience. For both of us.
Nothing helped until scuba. Scuba made microservices at scale *possible*. Not a world class team, not other monitoring or APM.
And learning how to *talk* about it. The differentiators, the positioning. Which industry changes were fueling the changing needs.
There is a 5-part blog where I first put forth and argued for a technical spec for observability, in fall 2017. It took me six weeks to write them. I did nothing else.
I failed at *everything* for a solid two years. It was numbing and endless. I cratered relationships. It was a bad time.
Maybe I can find other fuel besides 180-proof stubbornness and a religious experience. Maybe more sustainable.
Well, I'm trying to transmute what I actually feel into joy. 🥴. It actually angers me a little at how obvious and intuitive it all seems now.
I am not an optimistic person; hazards of the profession I guess. But those long, awful years of begging and pleading for users are receding fast, and that feels good.
We don't have the capacity to onboard everyone who wants to pay us lots of money.
Lots of signups come from people and companies we have never heard of.
And seeing just how quickly they, like us, understand that *there is no going back*. They take us with them when they switch jobs, because they cannot fathom engineering without it.
Thanks kids 🖤🐝