Checking Check All Routes

[SteveDMurphy]

Closes #2491

Code Changes

• Add logging where trying to align production build of routes

Steps to Confirm

• list any manual steps for reviewers to confirm the changes

Pre-Merge Checklist

• All CI Pipelines Succeeded
• Documentation:
  • documentation complete, PR opened in fidesdocs
  • documentation issue created in fidesdocs
• Issue Requirements are Met
• Relevant Follow-Up Issues Created
• Update CHANGELOG.md

Description Of Changes

This appears to only happen when using a production build of the front end, potentially due to the files being compiled in a different path than expected. Originally in 2.5.1

[SteveDMurphy]

Some example paths:

/usr/local/lib/python3.10/site-packages/fides/ui-build/static/admin
fides/src/fides/ui-build/static/admin
src/fides/src/fides/ui-build/static/admin
.venv/lib/python3.10/site-packages/fides/ui-build/static/admin

We may need to look at handling an infinite number of preceding path combinations and look at something like fides/ui-build/static/admin

[SteveDMurphy]

Took a few minutes to give this a little bit of a fresh look with what was discussed by @chriscalhoun1974, @sanders41, @Roger-Ethyca, and @seanpreston on the call tonight - it can definitely be improved with some decent coffee and some sleep but I think it might be a good start towards what we discussed! I was able to validate everything working as planned using nox -s "fides_env(test)"

[sanders41]

I was going to have another look also, but got unexpected results. I ran nox -s "fides_env(test)", went to a page with a route after the / running on 8080, refreshed and got no error. I thought it was weird because I didn't think we had it fixed so I tried on main, again no error.

I do still get errors when using 1.6.0 from docker with docker pull ethyca/fides:2.6.0 and pip install ethyca-fides, but can no longer reproduce it running from the repo. I verified there were no new commits on main from when we were able to reproduce it earlier, and also used incognito mode to rule out cache.

[sanders41]

I did some further testing by connecting to the 1.6.0 docker container pulled from docker hub and modifying files in the container with the changes here. I found the pages are loading with this change in place. I am getting a 500 error 2023-02-03 03:50:41.683 [WARNING]: Error sending analytics event: The fideslog API responded with an error: 500 Internal Server Error, but this seems unrelated and maybe the logging server is down?

I was initially concerned this would only work when using our docker image and not when building a custom docker image or pip installing fides. To test this idea I deleted the src/fides/ui-build directory, leaving the ui-build in the /usr/local package directory to mimic a pip install ethyca-fides. Doing this the pages still loaded successfully so it seems as though this is working. We discussed multiple ways to start fides, and I haven't tested them all yet so there is a chance there are more failure cases still to be found.

[SteveDMurphy]

I'll try and look at the fideslog stuff in a bit. Great catch! Even on failure we shouldn't surface anything if we can help it so am almost hoping it's something to do with the change...

[sanders41]

Just to clarify a bit. the 500 error shows in the server log, but pages on the front end still loaded fine without error.

[SteveDMurphy]

Ah cool, thank you!

[chriscalhoun1974]

Executed nox -s dev on this branch. Admin UI running on http://localhost:3000 works as expected when performing a page refresh via F5 on the top and side level navigation routes.

[chriscalhoun1974]

Just tested with nox -s "fides_env(dev)" and nox -s "fides_env(test)" commands. Page refresh is working as expected using http://localhost:3000.

[SteveDMurphy]

Closing in favor of #2500