Exactly, I completely reject the premise of this article. In the last hour I searched the "alerts" channel to see how many times our database had high CPU in the last year.
I'm not one to defend slack, but there's nothing inherently temporary about slack. And searchability > manual categorization.
That doesn't seem enough to reject the premise of the article, which explicitly calls out that chats are great for both "Low-stakes status updates" and "Swarming around red alerts or outages", a high CPU alert that was just a notification is the former, a high CPU alert that people discussed and fixed is the latter.
Also, your "database", singular? Was the answer something like 10 times? Would that scale?
I'm not one to defend slack, but there's nothing inherently temporary about slack. And searchability > manual categorization.