Today I was reading up on various technologies that allow a server attached to shared storage to handle high amounts of I/O locally. These technologies will “eliminate” the performance bottlenecks that systems encounter when shared resources are overwhelmed by today’s modern computing solutions.
About 6 years ago I remember reading how centralized storage solutions allowed a server to handle higher amounts of I/O and to perform better by leveraging the higher quality components found within SANs. That technology would eliminate the performance bottlenecks that local storage resources were overwhelmed by with the modern computing solutions of the time.
And long before that I am sure that someone was selling the benefits of local RAID arrays which would enable a standalone system to avoid the bottlenecks which came with shared mainframe resources. That technology would eliminate the performance bottlenecks that shared storage and resources were overwhelmed by.
I am not suggesting that these claims were false when they were made. Technology is always changing, and for every solution that we deploy we inadvertently create a new problem. In economics this phenomenon is referred to as the “law of unintended consequences”. In IT we call it progress.
Keep this in mind that next time you deploy a new technology. Do not tell yourself “I solved that problem for good!”
Instead start asking the question “What new problems have I introduced into the environment?”
If you take that approach you will be ahead of the problems that eventually arise in every environment no matter how good the technology is.