Move integration tests to 2.0

[FionnD]

Describe the feature
As we move towards 2.0, we need to redo each of our current integration tests and update them to the new codebase.

Integration tests in 1.x

When moving tests from issues to somewhere else, keep the issue ID in the docstring of the test.

TO WORK ON THIS

1. Put your name on a task you pick
2. Open a PR
3. Add the PR next to the checkbox
4. Tick the checkbox once the PR is done

(check first item below)

• crud. Basic: test: add basic crud integration #2501 cristianmtr
  • test workspace usage in CRUD (serialization) test: serialize crud and assert chunks #2521
  • rename to just crud test: serialize crud and assert chunks #2521
  • test chunks in CRUD test: serialize crud and assert chunks #2521
• rest usage in integration/flow cristianmtr test: crud rest #2525
• crud_corrupted_docs
  • to be removed
  • mime type detection under tests/docs cristianmtr test: add asserts for mime types #2531
• [ ] debug -- to be moved to hub
• doccache -- only on 'r' level feat: add doccache executor-indexers#21
• dump -- moved to hub integration refactor: CI, integration tests, cleanup executor-indexers#8 cristianmtr
• eval_flow. To be kept. Tests hanging pods for evaluation (name has changed: test_inspect_pods_flow.py)
• evaluation.
  • one part move to Hub, about the Evaluator @JoanFM
  • groundtruth test in core, check to see if you can pass GroundTruths to inputs test: added groundtruths routing #2520 maximilianwerk
• [ ] executors_batching. We don't have batching anymore. Responsibility of Executors
• [ ] executors_generic. Tested in other places.
• gateway_non_blocking -- added
• high_order_matches -- added
• hub_usage -- added
• incremental_indexing -- to be added to DocCache in Hub feat: add doccache executor-indexers#21
• issues
  • [ ] github_1072
  • check if filter is tested on DA cristianmtr test: docarray filter #2535
  • check if parameters are passed to f.post test: added groundtruths routing #2520 maximilianwerk
  • [ ] github_1229
  • check if docs_matrix is tested in Tail test: added a test for docmatrix in the tail pea #2537 maximilianwerk beware: Two test cases needs complicated fixing!
  • move MatchMerger to hub with travers path param -- added to https://github.com/jina-ai/internal-tasks/issues/92
  • github_1468 -- to investigate test: added a test for docmatrix in the tail pea #2537
  • [ ] github_1545 -- unnecessary
  • github_1546. To be added. Tests order of imports in pymodule test: added pymodules import test back #2536 maximilianwerk
  • github_1684 -- to be added to NpIndexer in hub test: added empty shard integration test executor-indexers#10 maximilianwerk
  • github_1861 -- test that if an exception is raised in init everything is closed properly test: added test for startup exception #2532 maximilianwerk
  • github_2103
  • github_2254 -- test that siblings from chunks are updated. Right now they need to be set manually. Might be remove and moved to tags. Siblings used to be updated from .extend refactor: remove sibbling from protobuf #2543 maximilianwerk
  • [ ] github_2295. Serialization is Executor based
  • [ ] github_2320 -- to be added to Hub for BinaryPbIndexer adapted to test growth
  • github_867 -- expensive closing of Executor is not interrupted by Flow. To be moved to Flow tests cristianmtr test: slow shutdown #2545
  • [ ] github_929 -- Will not be kept, but we should add serialization to the stress test jacobowitz deepankarm also move to BinaryPbIndexer in hub
  • [ ] github_969 -- Removed. But maybe should be put in a museum?
  • [ ] github_976 Removed
• [ ] level_depth. This is covered by multires example
• logging -- added
• memmap -- Did only test, that the numpy indexer does not crash and logged memory usage. Nice as a blueprint for future memory logging, but otherwise useless. Will be deleted.
• merge_pass -- to be adapted. First test in core, other in hub under MatchMerger
• mime -- add more mime type assertions to tests/unit/types/document/test_converters.py test: add asserts for mime types #2531
• [ ] multimodal -- covered in test document multimodal
• optimizers. To be adapted test: added optimizer integration test back #2519 maximilianwerk
• ranker_training. To be moved to Hub @bwanglzu feat: add lightgbm ranker executor-lightgbm-ranker#1
• [ ] ref_indexer -- Potentially not implemented again
• rolling_update -- no actual data (double check again if more tests/as fsertions need to be added) cristianmtr test: check distinct executors in roll update #2546
• sharding -- to be moved to Hub @cristianmtr test: sharding and npfilecompound executor-indexers#13
• sparse_pipeline -- added

[cristianmtr]

Some comments:

crud -- I guess here we would be testing on dummy Executors in the core?

dump -- This is there, but it's the DocumentArray.save implementation. The plan is to have the hub module of PSQLIndexer be the one that implements the exports and imports. Then people needing those would need to import that module. https://jinaai.slack.com/archives/C018F60RBL5/p1622190316138700?thread_ts=1622188950.135200&cid=C018F60RBL5

[cristianmtr]

@CatStark can we close this one?