It's definitely more complicated than that, I've migrated containers over to new OSes and I'm familiar with how tough it can be getting all your tests to pass a

It's definitely more complicated than that, I've migrated containers over to new OSes and I'm familiar with how tough it can be getting all your tests to pass again. That being said, it's a solvable problem. It's hard to say how many services Twitter runs internally, but by the sounds of this article their internal tooling sounds pretty lean and mean. A team of 5-10 tenured engineers would probably have no problem switching off of RHEL before the year is over (much less by June). As much as I want to watch Twitter collapse under the weight of it's own incompetence, it seems highly unlikely that Elon Musk is going to allow Twitter to go offline. If Twitter really was so fragile that it couldn't be scaled by throwing more money at the problem, then I guess I get why all the previous engineers were fired.