>Bryan Cantrill has compared the value proposition of an Oxide (rack?) to an IBM AS/400.

I've heard Bryan and Co. call it a "mainframe for Zoomers," but it's much closer to what Nutanix or VxRail is/was doing than it is to an AS/400.

It's not really a mainframe because the RAS story (Reliability, Availability, Servicing) story is sorely lacking compared to what a true mainframe gives you. So a midrange machine like AS/400 is probably a better comparison.

An AS/400 has a similar RAS story to mainframes than to Oxide/Dell. Oxide is closer to Dell (Oxide RAS is effectively the same as any sled hyperconverged) than they want to admit.

For modern applications written to seamlessly handle node failures the reliability is perfectly adequate.

For those of us who are unaware of "the value proposition" of an "IBM AS/400," could someone spell it out for us?

When the AS/400 came out circa 1989 or whatever, you could replace an entire mainframe with a box not much bigger than a mini fridge. The hardware is built for high reliability, and the OS and application software stack have a lot of integration. If Unix is "everything is a file" then AS/400 is "everything is a persistent object in a flat 64 bit address space."

The result is a system that can handle years of operation with no downtime. The platform got very popular with huge retailers for this reason.

Then in later years the platform got the ability to run Linux or Windows VMs, so that they could benefit from the reliability features.

High capacity, super reliable box that you could run your entire business stack on, if you could afford it.

The money IBM made with the AS/400 is actually completely mind blowing when you compare it to the rest of the computing industry at the time.