Make the timeout configurable for acquiring the DistributedLock in ContinueRunningWorkflows

[gersti07]

Currently in a multi node environment, a running Workflow on one node can block the ResumeWorkflow Startup on another node.

This PR adds an additional configuration setting to be able to have a custom timeout for acquiring the DistributedLock in the ContinueRunningWorkflows Startup, which resumes all running or idling Workflows.

With this setting i can define a short Timeout like 5-10s before i stop acquiring the lock.

If i can not acquire the lock in a short timespan, i assume that the other node is running this workflow, and i can skip resuming this workflow.
This design would potentially leave 2 edgecases unhandled:

• the lock acquisistion is very slow for different reasons (network problems, unresponsive DB or similar)
• Potentially: another instance hasn't correctly disposed its DB Lock

I choose the default timeout of 1h like before on the DistributedLockTimeout, to not break stuff, but maybe a default timeout of 5-10 seconds would be good.

see also: #4654

[sfmskywalker]

That all makes sense, thank you!