This is a story of spinning rust, and wading around, lost in the IO wilderness. I threw that wilderness thing in there for effect.
As a consultant at a hospital, I was fortunate enough to be in the loop on their first foray into the SAN storage space. Major RFPs and much sorting of contenders. At the time, I had already had a SAN implementation under my belt. A DEC StorageWorks array that was UltraSCSI based, the immediate forerunner to Fibre Channel arrays. IO tracing and troubleshooting has always been a passion. So much so, a network manager at that hospital would poke fun at me with a snarky comment: “It’s all about spindles.” Where my retort was: “It’s all about IO, Bob.” Having left there and taken a job at a large med-lab testing company, a co-worker and I would debate storage and IO. It got heated as he went on lecturing me about disk drive “spiral read rates.” That told me he was stuck in the early-90s as tagged-command queueing (TCQ or TQC) was already well entrenched and focusing on spiral reads was no longer a thing. Fast forward to joining a consultancy and my good friend Bergy (he being a top AE) would trot me into client sites. At first, it was a bit of stand-up that “Doctor IO” was there to discuss and advise. Doctor IO morphed into “The Good Doctor IO”. It doesn’t have to be painful to interact with customers and have fun solutioning. Dem were da days, right Bergy?