We all are aware of the hundreds of articles at blogs raving about how the performance in / around SSD is going to transform the storage industry. I am not going down that rabbit hole; I am however, going to present another angle all together. Have you considered how many SSD’s your disk subsystem can support without bringing it (disk subsystem) to its virtual knee’s? You must understand, it is one thing to support SSD’s (say as a RFP checklist response and it is quite another to have them operate in your Tiered ILM with out negatively impacting disk subsystem performance).
Have you considered how did the disk subsystem architecture changed to support the SSD’s? Was the disk subsystem powerful enough to support SSD’s before their implementation? What is an optimal utilization to run a disk subsystem at with SSD’s? What is an optimal SSD utilization rate (is it 30%, is it 50% or >)? What application will you run on SSD? What business unit will want to run their application on SSD’s (besides all of them)?
What critical thinking have you been through to get to the nuts and bolts of how this technology will impact your enterprise?