Fix tests pool failing

[AurelienFT]

• document all added functions
• try in sandbox /simulation/labnet
• unit tests on the added/changed features
  • make tests compile
  • make tests pass
• add logs allowing easy debugging in case the changes caused problems
• if the API has changed, update the API specification

The tests were failing because add_operations is managed in an other thread and so the tests don't wait that the ops are inserted before trying to fetch them. So sometimes it was fast enough sometimes not. With the new sleep it's always fast enough.

The test_pool failed after this change because it was wrong by design. Fixed it.

Fix #3509 and #3584

[Ben-PH]

LGTM. I'm gonna try something locally before I approve, but the essence of it is something like this:

for ((i = 0 ; i < 100 ; i++)); do
	cargo test --release #argument only for these tests here
done

[AurelienFT]

I will fix the tests that are not related to this PR but as it's a test fixing PR let's make everything here.

[Ben-PH]

I'm still getting some failures, but they are much less common:

for ((i = 0 ; i < 100 ; i++))  ; do
    cargo test test_pool -F testing --release
done

error output:

---- tests::operation_pool_tests::test_pool stdout ----
thread 'tests::operation_pool_tests::test_pool' panicked at 'assertion failed: ids.iter().map(|id|\n            (*id,\n                storage.read_operations().get(id).unwrap().serialized_data.clone())).eq(thread_tx_lists[target_slot.thread\n                            as\n                            usize].iter().filter(|(_, r)|\n                    r.contains(&target_slot.period)).take(max_count).map(|(op,\n                _)| (op.id, op.serialized_data.clone())))', massa-pool-worker/src/tests/operation_pool_tests.rs:166:21
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

improvement today is better than perfect tomorrow though. Mention this PR in the parent issue, then merge (subject to #3602 (comment))

[Ben-PH]

Would it be constructive to add a comment to one of the sleeps that provides this context (not a merge-blocking suggestion)

The tests were failing because add_operations is managed in an other thread and so the tests don't wait that the ops are inserted before trying to fetch them. So sometimes it was fast enough sometimes not. With the new sleep it's always fast enough.

[AurelienFT]

It shouldn't fail anymore will check on that also

[AurelienFT]

@Ben-PH Can you re-review I have fixed all the tests. I don't have any fail now.

[Ben-PH]

Nice. Those were bothering me, and didn't have much clue as to how to fix.

Just apply the same comment change "Allow some time for...", and for CI to finish and merge it.