I’ve been following the latest development updates over the last few months and I’ve been interested in the VPP Addon for some future plans in my home lab.
It’s clear to me that is still under development and experimental.
However, what I would like to understand is if the VPP Addon will also be Open Source eventually, or behind a paywall?
Hopefully some Sentrium (the company behind VyOS) can make a public statement in this thread regarding that toptic shortly.
In the meantime I sure hope VPP wont be placed behind a paywall which the “third party addons” would be a first step towards as I interpret whats written over at VyOS and VPP - progress and plans
Well, from the blog I was left divided, because in a way, it feels to be steps towards a similar system as Red Hat, where they have some stuff Open Source but than started to deviate into several paid solutions.
I do agree that with the future possibility of Third Party addons, putting the official VPP behind a paywall, but than allowing third party ones could result in duplicated addons and escalate possibly into another “vyatta fork” if people don’t like the future expectations.
In the end, the main reason I was left wondering this was because of the decision to be “opt-in” to test and get it early (which is fine to be honest), however without the code being available anywhere (I could be looking in the wrong place).
It is low power when even having 1 or 2 CPU cores pegged for VPP, the CPU itself still has a power limit.
So, yes, even using VPP, it will be more power efficient than having the equivalent required CPU power in a kernel base network stack.
Purpose of DPDK aka VPP is to better utilize multicore CPU’s which are available.
This way the same physical hardware can push more packets per second meaning higher throughput and lower latency compared to running without DPDK (VPP).
Promises are all good and well, however many companies and organizations in the past have broken them when the big paycheck shows up.
Still, these are promises regarding VyOS itself, the core product lets call it and doesn’t cover where VyOS is heading and its Addon capabilities.
I do consider that Addons will have the potential to make VyOS even better than it already is as platform.
However, VPP (or DPDK) is in my eyes and the eyes of many, the future of network processing, as we require more and more speed, as such being such an important feature, the stance from Syncer is very concerning, since avoid answering this question, is in a way an answer.
Leaves a lot to interpretation and makes you wonder where the mindset of the team is at.
All this on top of Syncer wanting to have VyOS ready for its “first funding round”, means he will want to make it profitable, as companies or entities won’t fund something if they can’t get something out of it.
Because of these 2 aspects, is why I was left concerned regarding all of these aspects.
But sure, let’s give it the benefit of the doubt and wait for an official answer from Sentrium.
I assume nobody needs 10G at home. But nobody needs a homelab either.
Homelabs are usually private playgrounds run as a hobby where people gain experiences.
Experiences in operating a Linux server, a hypervisor, running a network or even running a high throughput network which definitely has it’s own quirks.
Many of those people also work in tech and bringing those experiences into companies.
So if you ask from a commercial point of view, there is definitely a advantage in having tech available to people so they can use it, get a feeling for it and bring those experiences to potential customers.
Other than that, there is no need.
My homelab even has two 100G switches, has a 10G Uplink, a own ASN and is connected to various peering points.
I don’t need any of that to watch youtube, but it is fun to play around with, or just to test new ideas and methods. It would be definitely the first place where I would test VPP.