Spine Toolbox changes related to ditching Dagster

[ptsavol]

As per title.

To test Spine Toolbox without dagster, checkout branch issue_2523, make a fresh conda env and do pip install -r requirements.txt.

Re Issue #2523

Checklist before merging

• Documentation is up-to-date
• Release notes have been updated
• Unit tests have been added/updated accordingly
• Code has been formatted by black
• Unit tests pass

[codecov]

Codecov Report

Attention: Patch coverage is 60.00000% with 2 lines in your changes are missing coverage. Please review.

Project coverage is 65.75%. Comparing base (0c09a65) to head (0585c03).
Report is 4 commits behind head on 0.8-dev.

❗ Current head 0585c03 differs from pull request most recent head 23e7b5c. Consider uploading reports for the commit 23e7b5c to get more accurate results

Files	Patch %	Lines
spinetoolbox/headless.py	0.00%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##           0.8-dev    #2715   +/-   ##
========================================
  Coverage    65.75%   65.75%           
========================================
  Files          196      196           
  Lines        35869    35870    +1     
  Branches      6094     6094           
========================================
+ Hits         23584    23586    +2     
+ Misses       11023    11021    -2     
- Partials      1262     1263    +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ptsavol]

After a small edit, I'm not finding any problems here anymore. If anybody wants to check this out, it would be much appreciated. No need to check code if you're too busy, just execute a couple of projects and see if something doesn't look right. See install instructions above.

[soininen]

Stopping a Tool (tested with SpineOpt) does not seem to work. The hammer just keeps hammering and when I quit Toolbox, there is a rogue process that prevents Toolbox from shutting down properly.

[soininen]

Besides the above problem with stopping a workflow, I did not find anything of note. Nice work @manuelma and @PekkaSavolainen!

[manuelma]

Stopping a Tool (tested with SpineOpt) does not seem to work. The hammer just keeps hammering and when I quit Toolbox, there is a rogue process that prevents Toolbox from shutting down properly.

Does this work a lot better with dagster?

[soininen]

Does this work a lot better with dagster?

Yes. I didn't run into similar problem with 0.8-dev, though it isn't perfect, either.

[ptsavol]

Stopping execution depends on luck on 0.8-dev as well. Other times it works well but I just got this traceback after stopping SpineOpt in 0.8-dev branch.

Exception in thread Thread-30 (_wait_ping):
Traceback (most recent call last):
  File "C:\Python310\lib\threading.py", line 1016, in _bootstrap_inner
    self.run()
  File "C:\Python310\lib\threading.py", line 953, in run
    self._target(*self._args, **self._kwargs)
  File "C:\data\GIT\SPINEENGINE\spine_engine\execution_managers\persistent_execution_manager.py", line 319, in _wait_ping
    data = conn.recv(1024)
ConnectionResetError: [WinError 10054] An existing connection was forcibly closed by the remote host

And there was a rogue python process after trying to quit toolbox that I needed to end in the task manager to actually close the app.

[ptsavol]

Ok I see the problem, Clicking Stop in the jumpster branch does actually interrupt the console successfully but this does not seem to be signaled over to the toolbox side, which causes the hammer to keep on hammering. I'll take a look.

[soininen]

We got the hammer problem fixed with @PekkaSavolainen. I believe we are now on par with the previous Dagster implementation.

[ptsavol]

I'll update the changelog and run black. Do you mind if I merge when I'm done?

[soininen]

Do you mind if I merge when I'm done?

I suppose this is all the testing we're ever getting from this branch, so no, I do not mind.