I think a big risk is a cpu level security issue similar to meltdown or spectre that ends up weakening the hardware isolation between tenants to the point where it can be exploited on mass on the cloud providers to wreak havoc. The probability of something like this happening is very low but not zero, I would say same level of probability as datacenter fire or earthquake banks should be planning for how to handle this type of event.
I don't see that this risk is any different than a similar apocolyptic failure happening to your on-prem equipment. There's not much you can do about it differently than the cloud just add some extra controls and hope for the best.
I very much doubt that anyone would not use the cloud because of a theoretical de-isolation bug.
Also, by the time you found out, it would probably already be too late anyway if you were a victim. If not, you just switch it off.
I am not saying that they should not use cloud just that it is important to have a plan in place to deal with a unlikely but high impact security event affecting a cloud provider. Just like companies have business continuity plans in case a data center disaster they need to have plans for evacuate a cloud provider should they need too.
I put on my seat belt when I drive on the highways even though a nasty crash at 120 kph would likely kill me. Not using a seat belt because you will be severely injured anyway is not wise.
Given the amount of profit banks make what is the Downside of having them be resilient against public cloud failures?
We can be almost certain that there are sw and hw vulnerabilities that can be so exploited, given the rate of discovery and knowing what now-public hypervisor and cpu vulns a time traveler from today could exploit eg 5 or 10 years in the past.
Right now VM's are the better choice for Statefull use cases like running a a conventional databases that expects a real filesystem. In a few years I think container and container schedulers will get good at doing persistent volumes.
In the meantime I think a less known but great solution for reliabily creating
VM' is OpenSoure BOSH http://bosh.io/ is an excellent tool that will allow you to deploy VM's on most major IaaS including AWS, Azure, OpenStack, vSphere and others.
BOSH is a hard to learn and it has a different philosophy than typical configuration management tools such as chef/puppet/ansible ... etc but it is totally worth it once you have it you have an amazing power tool at your disposal.
I think the difference between an Airbus Airplane and a Software System is that once Airbus chooses a material that it meets it characteristics it does not go back an change that material. And the material does not change by itself. So the airbus test pyramid works because the bottom is stable.
In my personal experience in software systems is that the bottom is not stable you never end up selecting a material and having component that never changes after you build it. All layers of an application tend to change as the application changes and user needs change. Therefore you can't really say that oh we tested the bottom of the pyramid and know for sure that it works.
Writing end to end tests for application is quite hard work and requires a lot of though to design an application that can tested both at the unit level, the component level, and the system level.
Even though end to end testing is very hard it's value is massive as an industry we should be focused on lowering the cost of end to end testing rather than saying that unit is good enough.
With the old google maps I could count on the fact that I could hope on Google maps 60 seconds before I needed to head out find the directions I needed hit print and be on my way. Or I could have some one on the phone talking to me and saying something hey I am lost I am kinda in this area of town I could pump that into google maps and give walking directions or even driving directions realtime over the phone.
The new google maps is frustrating to use. All the ways that I used to use google maps for just don't seem to work with the new maps either because they removed the feature, or moved the UI elements in such a way that they don't exist on the same page any more or its just too slow on my very fast laptop and my very fast internet.
I really hope they fix it fast, I am loosing all hope that maps will ever be usable again.
