Gitactions improvements

[Polber]

This PR provides several improvements to the existing gitactions framework

• Modularized workflows
  • There was a lot of repeated code that is now modularized to reduce duplication and provide single place to define versioning of dependency actions (other than actions/checkout which is needed in each job to make sure code is up-to-date)
• Updated dependencies
  • Numerous libraries were throwing deprecation warnings, so those have been updated
• Added scheduled workflows for Unit Tests, IT's, Load Tests and Smoke Tests
  • These workflows are scheduled to run every 6 hours, and there will be status badges added to main page to track status of the various workflows
  • They will also be able to be run as dispatch jobs using a branch passed as an input
• All workflows not run against a PR (release and scheduled workflows) will now have a site report published as an aritfact
  • The action currently used to publish a test report only works for PR runs, resulting in sifting through messy logs for release failures. The site report will now be published as an artifact and can be downloaded and run in a browser to get an organized view of the test failure(s)
• Various other cleanup to improve organization and runtime of the actions

All workflows tested on personal fork

• Release: https://github.com/Polber/DataflowTemplates/actions/runs/8899421405
• Java Build: https://github.com/Polber/DataflowTemplates/actions/runs/8899925234
• Java Unit Tests: https://github.com/Polber/DataflowTemplates/actions/runs/8899927780
• Java Integration Tests: https://github.com/Polber/DataflowTemplates/actions/runs/8899926003
• Java Smoke Tests: https://github.com/Polber/DataflowTemplates/actions/runs/8899926914
• Java PR: https://github.com/Polber/DataflowTemplates/actions/runs/8899944601 (With failing tests)
• Java PR: https://github.com/Polber/DataflowTemplates/actions/runs/8900110084 (With successful tests)

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 0.00%. Comparing base (b7f8c2b) to head (4e4b304).

:exclamation: Current head 4e4b304 differs from pull request most recent head fa87c37. Consider uploading reports for the commit fa87c37 to get more accurate results

Additional details and impacted files

@@             Coverage Diff              @@
##               main   #1495       +/-   ##
============================================
- Coverage     40.45%   0.00%   -40.46%     
============================================
  Files           732     250      -482     
  Lines         41628   13789    -27839     
  Branches       4475    1477     -2998     
============================================
- Hits          16842       0    -16842     
+ Misses        23319   13789     -9530     
+ Partials       1467       0     -1467     

Components	Coverage Δ
spanner-templates	0.00% <ø> (-56.11%)	:arrow_down:
spanner-import-export	∅ <ø> (∅)
spanner-live-forward-migration	0.00% <ø> (-61.24%)	:arrow_down:
spanner-live-reverse-replication	0.00% <ø> (-42.54%)	:arrow_down:
spanner-bulk-migration	0.00% <ø> (-70.56%)	:arrow_down:

see 629 files with indirect coverage changes

[damccorm]

Also this large PR contains many changes to the test infrastructure. I would suggest split into smaller PRs and merge one/several a time to reduce the distraction to the developments.

Yeah, I agree with this; I like the direction of this set of changes generally, but splitting into smaller prs will make it much easier to review/revert if certain components are broken at all (this is especially important since testing CI is hard and can't be done in an obvious automatic way)

[Polber]

Fair enough. Smaller PR's TBD. Thanks

[Polber]

Opened #1519, #1520, #1521, #1522 and #1523