try tweaking some times to reduce flakiness

[BinaryFissionGames]

Using this PR to try testing some changes to tests to reduce flakiness, since they don't seem to flake locally.

[codecov]

Codecov Report

Merging #558 (c97bd1a) into master (3b23e5c) will decrease coverage by 1.95%.
The diff coverage is 73.17%.

@@            Coverage Diff             @@
##           master     #558      +/-   ##
==========================================
- Coverage   76.69%   74.73%   -1.95%     
==========================================
  Files         132      132              
  Lines        8308     9993    +1685     
==========================================
+ Hits         6371     7468    +1097     
- Misses       1473     2056     +583     
- Partials      464      469       +5     

Impacted Files	Coverage Δ
operator/buffer/buffer.go	75.86% <ø> (-1.92%)	⬇️
operator/buffer/memory.go	71.60% <ø> (-3.74%)	⬇️
...builtin/input/aws/cloudwatch/cloudwatch_persist.go	75.00% <ø> (-6.82%)	⬇️
operator/builtin/input/azure/event_hub.go	0.00% <0.00%> (ø)
operator/builtin/input/azure/event_hub_config.go	90.91% <ø> (-1.09%)	⬇️
operator/builtin/input/file/finder.go	100.00% <ø> (ø)
operator/builtin/input/file/reader.go	53.05% <0.00%> (-5.73%)	⬇️
operator/builtin/input/http/config.go	97.56% <ø> (-0.29%)	⬇️
operator/builtin/input/journald/journald.go	58.76% <ø> (-1.83%)	⬇️
operator/builtin/input/udp/udp.go	73.49% <ø> (-1.51%)	⬇️
... and 156 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update cd2849f...c97bd1a. Read the comment docs.

[BinaryFissionGames]

Ran 5 successful tests in a row; Empirically, using the tests cases exactly as the are from https://github.com/open-telemetry/opentelemetry-log-collection seems much more stable.

It's not 100% transparent to me why it's more stable, however.

[cpheps]

The fact that just tweaking time values makes it work is a bit of a code smell but I'm ok with this being a temporary mitigation.

[BinaryFissionGames]

https://github.com/observIQ/stanza/runs/5249699444 Another failure, on windows of course. I'll come back to this in a bit.

[BinaryFissionGames]

@cpheps I looked into it a bit closer since I got some test failures locally at some point (some tests will actually deadlock if they fail by not polling a log entry fast enough, which confused me), and yeah, it's just generally hard to test these things better with how the file input operator is right now.

A future goal might be refactoring a bit and have this better testable, and rewriting some of these tests so they aren't so reliant on timeouts.