Managing networks and storage in a contracting economy may get difficult. Will CFO's decide that (4) 9's reliability is more cost effective than (5) 9's reliability? Perhaps they will decide that 99% (3.65day/year) is good enough for networks and storage.
When it comes down to reliability of a multitude of components, people , Internet , electrical grid, and weather what is a realistic expectation of service from your network and storage infrastructure? And how can you put a value on up time when there are so many variables ?
From the discussions I have had this week, it seems that many people are looking at their networks and storage infrastructures and trying to define what is a realistic up time guarantee. I really don't think anyone thinks that data loss is acceptable, but the question did come up in one meeting of who is going to take care of a bankrupt company's data to make sure it is reliable. In a paperless world how will the next generation look back to see what happened, who will keep the architectural drawings and piping diagrams of chemical plants for example when the companies that did the drawings and their customers are gone, and the disks are unavailable?