Iwan Budi Kusnanto

Hi @scottyeager @LeeSmet I can reproduce it quite easily and can see the difference using `ping` i did three experiments, each of them only use one public node to simplify...

It would be good if we can do profiling is this the recommended way @LeeSmet ? https://github.com/threefoldtech/mycelium/issues/141#issuecomment-2076641443

I did a bit check for the data paths: 1. I remember that there was a plan on doing TUN offload #141 . @LeeSmet what is the state of TUN...

> The TUN offload issue has been parked as its purely an optimization for the send/receive nodes. It can be done in the future once the system is completely stable....

> We could use tokio's channel recv_many on linux followed by a write vectored for some performance improvements for now, that should be a quick and easy fix I've tried...

> the most obvious thing is because my system is not busy enough, there is small chance that there are more than one message that ready to be consumed, so...

FYI, i discarded the work on TUN. It was started because i misunderstood how TUN device being used in the `mycelium`. Because Lee works on the routing part, i'm checking...

> Error is returned immediately if the kernel param is set. > So I think deployment to a node with disable-gpu is always gonna fail. nice catch 👍

> Mycelium worked fine on Windows > but could not do the same to machines deployed from the grid from windows Hi @khaledyoussef24 Could you try to change Windows with...

OK, i can reproduce it, using command line mycelium, not the flutter one. Did few scenarios 1. windows to Linux(my PC) - ping OK - ssh OK 2.Linux (my PC)...