If I can summary from above statements, what makes Nimble`s Active/Passive design is unique/different compared to another storage vendors design.
Where different active/passive designs differ, is the time to fail over is even longer than non-symmetric active/active designs. A VNXe can take over a minutes, a Dell EqualLogic can take 35 seconds. If this is longer than an instant you risk IO timeouts on applications; delayed write failures; NTFS corruption; EXT4 going read only; and generally angry users.
I've never tested a Nimble for failover; I'm curious how long it will take.
John is spot on for Active/Passive designs...they can very seriously affect failover times and thus data access. This is why we refer to the Nimble architecture as Active/Standby. The Standby controller is always up and running (it's actually doing a few minor things in the background) it's just not processing host I/O load. In case of failover it's read to take on the load seamlessly without any loss to data access or prolonged failover windows. You will typically only see 2-3 ping drops on the data IPs if you fail the Active controllers. The official answer for the design is as John pointed out: guaranteed SLAs upon failover. If you run active/active systems with a best practice design of being able to handle failure of a controller, you would still run no more than 100% 'system' load, which is what we do on a single controller at all times. If you remember the Yahoo! Fantasy Football outage, a main reason for that was overloading the systems controllers on a failure and bringing down the whole system. An Active/Standby configuration allows us to guarantee that can't happen. There's nothing that limits us from running Active/Active, it was just a very conscious design goal. And yes, we also did happen to build a REALLY col filesystem in the process :).
Hit the app-data gap? HPE Nimble Storage flash arrays and solutions with predictive analytics ensure fast, reliable access to data for data center and cloud applications. This allows you to eliminate the gap and guarantee 99.9999% availability with the best all-flash capacity per TB in the industry.