fix: single-node lighthouse startup issue

[Dyslex7c]

Closes #1053

[barnabasbusa]

Thats just fixing it for one example.
We want a fix that will fix it for each run where there is only a single lh participant.

[Dyslex7c]

Thats just fixing it for one example. We want a fix that will fix it for each run where there is only a single lh participant.

got it, will update this PR soon

[Dyslex7c]

I guess the --target-peers=0 parameter is enough to fix this problem, although this might cause issues for multiple participants (> 2) @barnabasbusa please do check my latest commit and correct me if I'm wrong

[barnabasbusa]

You need to have it as an optional flag set under lighthouse, depending on how many participants each run has.

[Dyslex7c]

@barnabasbusa I've added the required condition in the starlark file, do let me know if there are any more changes to be made

[barnabasbusa]

Your PR doesn't seem to be working.

[barnabasbusa]

Please only convert the pr back to open if you have actually tested your own pr.

[Dyslex7c]

@barnabasbusa I tested my PR with both a single participant and two participants and it seems to be working. Here are the screenshots based on which I concluded

[barnabasbusa]

its not appending the target-peers=0 with 1 participant only.

You can inspect the running container or run with --verbosity detailed

[Dyslex7c]

@barnabasbusa I moved the logic to main.star from participant_network.star for debugging and it seems to work, don't know what went wrong. Can I keep the logic inside main.star? I guess there won't be a problem since I'm checking the cl_type anyways

[Dyslex7c]

@barnabasbusa I've made changes and tested the PR, it is now appending --target-peers=0 for just one lighthouse participant

[mininny]

@barnabasbusa I think this PR breaks existing usecases where we create only one client pair in kurtosis and connect it to a running devnet/testnet (i.e. kurtosis-sync-text). Do you think this change makes sense to fix this? #1119