[Feat] Add graceful detection of signal to exit + SignalConnector and merge SlurmConnector.

[tchaton]

What does this PR do?

Right now, when the script is being killed, a fault tolerant checkpoint would be saved.
However, Lightning can't always ensure it is always fully happening in reproducible part of the codebase.

When running in the cloud, there is 2-3 min to kill the script.

The proposal is to add a mechanism to detect a killing signal as been sent and Lightning will terminate on the next reproducible part in this time windown.

Parts #9567

Does your PR introduce any breaking changes? If yes, please list them.

Before submitting

• Was this discussed/approved via a GitHub issue? (not for typos and docs)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure your PR does only one thing, instead of bundling different changes together?
• Did you make sure to update the documentation with your changes? (if necessary)
• Did you write any new necessary tests? (not for typos and docs)
• Did you verify new and existing tests pass locally with your changes?
• Did you list all the breaking changes introduced by this pull request?
• Did you update the CHANGELOG? (not for typos, docs, test updates, or internal minor changes/refactorings)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:

• Is this pull request ready for review? (if not, please submit in draft mode)
• Check that all items from Before submitting are resolved
• Make sure the title is self-explanatory and the description concisely explains the PR
• Add labels and milestones (and optionally projects) to the PR so it can be classified

Did you have fun?

Make sure you had fun coding 🙃

[github-actions]

Build Error! No Linked Issue found. Please link an issue or mention it in the body using #<issue_id>

[codecov]

Codecov Report

Merging #9566 (7216a39) into master (982a956) will decrease coverage by 4%.
The diff coverage is 71%.

@@           Coverage Diff           @@
##           master   #9566    +/-   ##
=======================================
- Coverage      93%     89%    -4%     
=======================================
  Files         180     181     +1     
  Lines       15094   15341   +247     
=======================================
- Hits        14021   13623   -398     
- Misses       1073    1718   +645     

[ananthsub]

does signal handling generalize across all clusters/schedulers?

our scheduler offers a dedicated mechanism to check for preemption, which we've integrated in the model checkpoint callback on batch end for whether we ought to save a checkpoint that batch. but i also thought this was atypical in cloud providers workloads where the job scheduler could kill the job at any point

[yifuwang]

Think looks great! Thanks for adding the mechanism @tchaton!

[awaelchli]

this line got unindented, it was previously under the slurm check.
it means that Lightning can't be killed by sigterm.
As best as I can judge, this is not required for fault tolerant training.

See #10154 for context.