This reverts commit d3bc298f.

julian.baeume authored 1 year ago and julian.baeume committed 1 year ago

Changelog:
- introduce two separate roles for middleware deployments:
    - ui-middleware: read from redis only, listen to version updates, can easily be scaled
    - updater: split out all code that writes to redis, no need to scale > 1

- Unit and Integration tests now use "inject" instead of starting a server
  and using supertest.
- Add Coverage Reports and "test:dev" mode. This enables a mocha --watch
  like workflow using nodemon, as --watch does not work with esm modules.
- Reorganized and renamed files and functions

Set cache control to no-store. The ui-middleware will then only look at the remote servers for the files and never in the local cache

Cache lookups will immediately register a promise to prevent multiple requests for the same ressource

Root cause: If multiple connections are opened at the same time after an update, the ui-middleware will do multiple requests to get initial data. That went especially worse, when e2e-tests are launched against ui-middlewares with multiple clients connecting at the same time.
Soltion: Make sure, that a node already caches stuff without a response.

Root cause: cache keys handled differently when reading/writing
Solution: make sure keys are handled equally

Fixes OXUIB-1528

Use redirect instead of delivering index.html to prevent /ui paths in the future

Root cause: There are middlewares, which still use old redirects to /ui
Solution: To prevent 404s, we just serve the index.html

Fixes #6

this is an implicit endpoint served by us.

first hit wins. This allows us to transparently use sourcemaps and other files that
are not part of the sources according to the vite manifest

this will mark the pod "ready", once the cache is warm.
This will remove the penality for the very first to this service, because
it will not receive any traffic until the cache is warm.

we want to have the changes rolled out, so we disable CSP headers for now and
will take care of them after the file_cache functionality has been rolled out

just for testing purposes, disable CSP for now until sha calculation is fixed

this currently doesn't really work as expected, but can be used as a base
to properly implement CSP for the scripts we "know" about.

the file cache is populated with data defined in the manifest files configured for
all components and will basically fetch everything from all components. This allows
us to use this manifest service as the only source of truth for UI related sources and
simplifies ingres setup a lot.