Add manual cluster restart options

[andrussal]

This extends StartNodeOptions in two ways:

• adds launch args, so local manual-cluster restart can exercise real CLI overrides such as --net-initial-peers
• adds runtime.start_timeout, so tests can override readiness timeout per node when needed

use std::time::Duration;

use testing_framework_core::scenario::StartNodeOptions;

cluster
    .restart_node_with(
        "node-2",
        StartNodeOptions::default()
            .with_args([
                "--net-initial-peers",
                "/ip4/127.0.0.1/udp/30001/quic-v1",
            ])
            .with_start_timeout(Duration::from_secs(180)),
    )
    .await?;

[hansieodendaal]

Question: Should PeerSelection not be compared to what it was before the restart?

Otherwise looking good.

[andrussal]

Yes this wasn't ideal, I made PeerSelection an Option like other fields so we don't have to handle it explicitly somehow.

[hansieodendaal]

Not directly related to this PR: Could we maybe add a configurable start timeout, instead of only using 60s?

[andrussal]

Not directly related to this PR: Could we maybe add a configurable start timeout, instead of only using 60s?

I added also option to override, see updated PR description.

[hansieodendaal]

LGTM!