They have a service 'infosite' that every array dumps stats to.. From there, they can predict things based on your usage patterns, such as when you are going to

They have a service 'infosite' that every array dumps stats to.. From there, they can predict things based on your usage patterns, such as when you are going to run out of disk space, when you should add additional SSD's (if a hybrid setup), when to upgrade controller models to handle increased IO, when a hard drive looks like it might need replaced, etc. It will also tell you what VM's have the most IO Load in vmware, how much bandwidth you need for replicating volumes, etc. My regional engineer told me they have one of the largest HP Vertica database clusters in the world to handle all the datapoints..