Test different resharing senarios and fix some bugs

[ailisp]

When test resharing I found some problem:

1. new participant not generating triples
2. as participant join and leave, some node can panick and crash

The root cause of 1. is the new participant has url with trailing /. And node ping ip:port//state (double slash) fails therefore not joining it into active participant set. Although the test infra can manage to spin up node with out trailing /, our node should be handle this case more robustly. The fix is to replace every string url concat with Url.join.

The root cause of 2. is as participant join and leave, stable participant set could be temprorailly below threshold number of participant. The fix is to wait and soon new active participant will become stable and protocol will proceed.

Also merge and improve resharing tests to test more senarios: leave, join, rejoin

[ailisp]

Seems the problem is more serious, log shows that new joined participants don't generate triplet at all. Still investigating

[ailisp]

Do you guys know why cargo check fails? There is no relevant change or ci toolchain change since last passing CI. They both use stable 1.79.0. Using stable 1.79.0 cargo check succeed locally. I found this https://users.rust-lang.org/t/time-crate-compilation-error/111789/3 but it only fails for a nightly rustc

[volovyks]

@ailisp We have this issue on Phuongs PR, the reason is unknown. It works on my PC and on the GCP VM.

[github-actions]

Terraform Feature Environment Destroy (dev-692)

Terraform Initialization ⚙️success

Terraform Destroy success

Show Destroy Plan

No changes. No objects need to be destroyed.

Either you have not created any objects yet or the existing objects were
already deleted outside of Terraform.

Destroy complete! Resources: 0 destroyed.

Pusher: @ChaoticTempest, Action: pull_request, Working Directory: ``, Workflow: Terraform Feature Env (Destroy)