On Mon, 30 Jan 2023 at 17:34, Ernst J. Oud <ernstoud@gmail.com> wrote:
Lukas,
Keep in mind that - as I experienced - any VM or Docker container introduces some problems, such as higher latency. My CentOS VM has 2 ms. higher latency in the first hop compared to a v5 probe on the same fiber.
Any virtualization may introduce variables, any WAN circuit, any Firewall and any NAT device. Whether the ATLAS SW probe runs as a "native install" on a virtualized VM managed by you, or whether the ATLAS SW probe runs in a VM managed by RIPE doesn't make a difference in your latency deviation example. We need to stomach 2 ms of deviation if we want to allow SW probes, the alternative is to go HW only.
And a bug in Docker causes traceroute to fail.
Docker is probably not the way to go, containers make it more easy to install, but I don't think a container can trigger its own upgrade.
So a hardware or a native install of a software probe is the best way to go.
I disagree, unmaintained and unsupported probe installation can never be the way to go. Lukas