Revert "conn pool: use hostnames of endpoints as SNI values (#34898)"

[phlax]

This reverts commit 6db316f.

Commit Message:
Additional Description:
Risk Level:
Testing:
Docs Changes:
Release Notes:
Platform Specific Features:
[Optional Runtime guard:]
[Optional Fixes #Issue]
[Optional Fixes commit #PR or SHA]
[Optional Deprecated:]
[Optional API Considerations:]

[repokitteh-read-only]

CC @envoyproxy/api-shepherds: Your approval is needed for changes made to (api/envoy/|docs/root/api-docs/).
envoyproxy/api-shepherds assignee is @markdroth
CC @envoyproxy/api-watchers: FYI only for changes made to (api/envoy/|docs/root/api-docs/).

🐱

Caused by: #35212 was opened by phlax.

see: more, trace.

[phlax]

cc @dmitriyilin - this introduced a frequent CI flake (see eg https://dev.azure.com/cncf/envoy/_build/results?buildId=175233&view=logs&j=8c169225-0ae8-53bd-947f-07cb81846cb5&t=d1a98671-b7ba-5fbf-f06c-ff337c010df4&l=230)

[dmitriyilin]

Thanks, I will take a look. However, i will be out of office for the rest of this working week.

As a notice, I've experienced flakiness with some other integration tests in checks in my PR. On top of that, a check in current state is also failing due to a failing integration test. Is there a known issue with stability of integration tests?

[phlax]

I've experienced flakiness with some other integration tests ... Is there a known issue with stability of integration tests?

its an ongoing challenge preventing, catching and fixing flakes - esp those caused by timeouts (in some runs/situations)

just as your PR got through by retesting it until the flakey test passed - so it is with other PRs - the issue/s introduced then affect other runs

while we try to be vigilant about this, its not easy - but we do give this more attention when we are preparing a release (as now)

there is a separate, (i believe) unrelated flake showing now

/retest