The observability/monitoring/debugability arguments are super old arguments that simply aren't true in today's (and most of yesterday's) unikernel world. Most of the current day observability tooling I see in use today like prometheus and such work out of the box.

Also, something without TCP support of which every single unikernel I know of has sounds extremely limiting to me.

Thanks. I'm admittedly out of touch. Now that I have your attention: What kind of startup times are we looking at with unikernels, say, to serve HTTP traffic with Nginx / Apache / Caddy or some such server in the front. I ask that because Workers start in 5ms or less: I see end-to-end response times as low as 25ms to complete the work my toy app's doing.

I'm genuinely interested in any low latency isolation tech that lets me build a multi-tenant solution. The point with Cloudflare Workers is, they've made it simple to deploy the code across all their PoPs. Are there Serverless / Unikernels at Edge providers that do something similar?

5ms happens to be a special number. NEC reports booting unikernels that fast:

http://cnp.neclab.eu/projects/lightvm/lightvm.pdf

If you're doing NFV type of workloads fast boot time might be interesting if you are booting up and shutting down things as fast as you can, however, if you are doing traditional cloud workloads ala jvm, rails, etc. I don't know if achieving fastest time to boot makes a ton of sense - at least for those app envs it takes a while just to start and would defeat the purpose of spinning up/down. Your example of web servers typically would indicate some form of db connection taking place - even if it's just a static webserver - why need to turn on/off so fast?

I'd be interested in hearing some specific usecases cause unikernels definitely hold promise here but I don't have a ton of use-cases outside of NFV.

The low latencies that typically get a lot of us excited are the low run-time latencies and high throughput promises unikernels have.

I'm interested in IPv4 in IPv6 and vice versa, TCP in UDP and so on for the data-plane. Imagine something like TailScale.com [0] but managed and not P2P, or maybe running something like Wireguard from within a unikernel. I'm not even sure if it is a valid use-case to have long-running unikernels? Also, does that count as NFV/SDN? Currently, I run the data-plane workload isolated via VMs and they are always-on, even when the traffic is non-existent.

The requirement for HTTP workloads is mostly for control-plane APIs, and imo, Workers will work really well, for that.

[0] https://tailscale.com