Historically, systems monitoring and management occurred in silos. Storage, server, network, and middleware teams each only monitored their slice of the stack, and in the best case (but not always), teams stitched together a performance assessment of the whole system after the fact.