Recently Steve Duplessie posted a video blog; he was inspired by sights during a trip to Italy to ponder the fact that much in IT doesn’t go away. That in turn (perhaps motivated by my lack of an Italian visit!?) made me think about the things in IT that we actually do want to get rid of, or at least reduce…and how those very things are the main reasons why some infrastructural approaches stay with us for the long term! Here’s my video commentary:
Steve Duplessie, ESG's esteemed founder, he is my boss, recently posted a video capsule that was engaging and insightful. Inspired by visits to Pompeii and Rome, Steve pointed out that things in IT often change in detail but don't go away in principle or even practice. Mainframes, servers, tape, etc. persist. Clouds, flash or containers won't obliterate everything that preceded them.
It reminded me of a balancing truism. The things in IT that we have always wanted to get rid of or at least reduce: cost, time, management effort, and risk. There's an important link here. Important because it's a reminder to focus on what's really needed rather than being distracted by shiny IT objects.
The reason the things Dupe talked about as not going away stay is not nostalgia or laziness or inertia, but it's simply because they continue to serve some of more of those other foundational desires. Some things do go away. Punch cards and fax machines are chariots that got replaced by mopeds and cars. But things that are good at optimizing or reducing resource usage and risk will stay.
So while private data centers might not optimize cost or even flexibility, they are perceived as helping to control risk. Meantime, clouds and containers will find their place and will stick around as long as they continue to help get rid of something. In their cases, it's some mix of time, cost, and effort that we do want to go away.