Better interrupts + node sorting

[Jorres]

Several quality of life changes:

• [bugfix] add transport-layer timeouts to avoid stuck grpc connections
• [bugfix] ctrl+c now interrupts correctly (previously not all points were interruptible and that led to delays depending on parameters, up to 30s with defaults)
• [no-config feature] an attempt to optimize per-tenant restart: assuming that CMS tries to give nodes in the order of submission, sorting round-robin by tenants gave a speedup (it was actually considered back when per tenant restart was implemented but due to human error didnt go into [NBYDBOPS-1669] add support for parallel tenant nodes restart #81 previous PR)
• [no-config feature] soft-prefer to restart nodes from a single failure domain (compared to just strict node-id sorting)
• new tests for all the above

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

[Copilot]

Pull request overview

Copilot reviewed 14 out of 14 changed files in this pull request and generated 6 comments.