REP-7721: Cleaning up sleeps in functional tests

[pglass]

This cleans up nearly all of the unnecessary sleeps in the functional tests.

• Sleeps that waited for repose to start or reload a config, or just to wait for some thing to happen, are replaced with log searches or other wait logic where possible
• A bunch of sleeps of 500ms were waiting for the repose client cache to clear. Since these are short, I kept the sleeps but moved them into a waitForHttpClientRequestCacheToClear that is more descriptive
• Various tests need to sleep and I left these alone, like the rate limit tests and various tests of socket timeouts and cache expiration

[pglass]

These tests were failing in Jenkins after replacing sleeps, so I reverted those changes.

• repose-aggregator/tests/functional-tests/src/integrationTest/groovy/features/filters/identityv3/cache/InvalidateV3CacheUsingAtomFeedTest.groovy
• repose-aggregator/tests/functional-tests/src/integrationTest/groovy/features/filters/identityv3/cache/InvalidateV3CacheUsingAuthenticatedFeedTest.groovy

[pglass]

Oh, actually the keystone v2 versions of those tests were failing as well. I'll wait for Jenkins to finish and revert those as well (curious if they still fail on a retest).

[pglass]

There's also a features.core.valveselfconfigure.RuntimeSysmodChangesTest that seems flaky with my new changes. It's failed twice in the past 20 runs or so, and both failures were on this PR: https://jenkins.openrepose.org/job/repose-gradle-pull_request-core/86/testReport/junit/features.core.valveselfconfigure/RuntimeSysmodChangesTest/history/

I'll revert changes for that test for now, and I'll make a note on the JIRA about the tests where my log searches didn't quite work out.

[pglass]

retest this please