xds: Add cluster endpoint watchers to depndency manager

[eshitachandwani]

This is part of A74 implementation which add CDS/EDS/DNS watchers to the dependency manager. It also adds a temporary flag that is disabled by default so that it is not used in the current RPC paths , but enabled in the dependency manager tests.

RELEASE NOTES: None

[codecov]

Codecov Report

❌ Patch coverage is 80.96591% with 67 lines in your changes missing coverage. Please review.
✅ Project coverage is 83.34%. Comparing base (6ed8acb) to head (cd5d990).
⚠️ Report is 2 commits behind head on master.

Files with missing lines	Patch %	Lines
internal/xds/xdsdepmgr/xds_dependency_manager.go	80.80%	50 Missing and 17 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #8744      +/-   ##
==========================================
- Coverage   83.39%   83.34%   -0.05%     
==========================================
  Files         419      418       -1     
  Lines       32566    32897     +331     
==========================================
+ Hits        27159    27419     +260     
- Misses       4023     4084      +61     
- Partials     1384     1394      +10     

Files with missing lines	Coverage Δ
internal/xds/resolver/xds_resolver.go	88.76% <100.00%> (+0.12%)	⬆️
internal/xds/xdsdepmgr/xds_dependency_manager.go	80.87% <80.80%> (-4.04%)	⬇️

... and 22 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[easwars]

Can you also please check the codecov report here: #8744 (comment) to see if there are any valid lines that are missing test coverage. Thanks.

[easwars]

Doesn't seem to be fixed yet.

[easwars]

Hasn't been fixed yet.

[easwars]

I made a small refactoring commit to split the populateClusterConfigLocked method into smaller ones to handle updates for specific cluster types.

I also added the default case for the switch at the end. Please note that I changed it return true, nil, nil instead of false, nil, nil as it was doing earlier, because in this case, we have received the cluster resource. It just happens to be a type that we don't support. That shouldn't block us from sending an update if everything else in the tree is resolved (since we are anyways storing the error in the cluster config).

[easwars]

Some of my comments might be outdated as this point, as I pushed a couple of commits for more generics and some simplification around the dns resolver.