[BUG] mgmt, investigate on occasional thread stuck

[XiaofeiCao]

We've had several customer reports of occasional thread stuck when using mgmt SDK.
Need to investigate on each one and maybe there's a shared root cause.

• [BUG] UndeliverableException when creating resource group and network security group in heavy load #33056
• [BUG] reactor.core.Exceptions$ReactiveException: java.lang.InterruptedException occurs at computeManager.virtualMachines().getByResourceGroup(resGroup, name) #33829
• [BUG] Azure Java SDK timeout requesting target resources #36808
• IcM, timeout exception: https://portal.microsofticm.com/imp/v3/incidents/details/391847100/home (turned out to be an external timeout)
• Tooling team reported thread stuck when calling convenience method, resolved by calling serviceClient method(spring app backend creating managedcontainer for quite some time and looks like stuck)

### Tasks
- [ ] Investigate `Schedulers.io()` on the behavior of thread shortage
- [ ] Test Completeable.complete() as return value for Completeable
- [x] Netty configuration
- [ ] Investigate on delete code where the polling stuck occurs