I use it to describe a variety of things, but usually it’s related to servers not being able to handle load rather than an outright crash, but I’m not strict about it. Laos balancer failures could be it, could also just be that something was really I efficient but wasn’t noticed until it went into production.