One serious doubt I still have is this: there’s a huge open-source community producing excellent building blocks, but turning those into integrated, production-grade edgecloud platforms seems to require either very large capital (i.e. big tech) or a level of sustained focus that’s hard to achieve.
That raises a question for me. Is the missing piece actually money and scale or is it where and with whom this integration happens? In other words, can domain-specific Edge-cloud systems realistically be built and maintained by smaller, focused teams working closely with the stakeholders who need them, or does this inevitably collapse back into hyperscaler-led platforms?
Anyway would an initiative such as now being launched in Europe, the OFEIA fly ?
https://www.linkedin.com/posts/open-federated-edgecloud-infr...
To be clear, I’m not arguing against hyperscalers they’re obviously extremely good at what they do. The question is whether all device-adjacent workloads should default to them, or whether there’s value in introducing an autonomous first hop that can enforce locality, policy, and failure containment before anything goes upstream.
On the security point: I’m genuinely interested in counter-examples. My intuition is that attack surface and attack impact are often conflated. A wider mesh may expose more entry points, but the blast radius of compromise is much smaller. Curious how people who’ve operated distributed systems see this trade-off in practice.
One serious doubt I still have is this: there’s a huge open-source community producing excellent building blocks, but turning those into integrated, production-grade edgecloud platforms seems to require either very large capital (i.e. big tech) or a level of sustained focus that’s hard to achieve. That raises a question for me. Is the missing piece actually money and scale or is it where and with whom this integration happens? In other words, can domain-specific Edge-cloud systems realistically be built and maintained by smaller, focused teams working closely with the stakeholders who need them, or does this inevitably collapse back into hyperscaler-led platforms? Anyway would an initiative such as now being launched in Europe, the OFEIA fly ? https://www.linkedin.com/posts/open-federated-edgecloud-infr...
To be clear, I’m not arguing against hyperscalers they’re obviously extremely good at what they do. The question is whether all device-adjacent workloads should default to them, or whether there’s value in introducing an autonomous first hop that can enforce locality, policy, and failure containment before anything goes upstream.
On the security point: I’m genuinely interested in counter-examples. My intuition is that attack surface and attack impact are often conflated. A wider mesh may expose more entry points, but the blast radius of compromise is much smaller. Curious how people who’ve operated distributed systems see this trade-off in practice.