Fix or evaluate impact of failing test cases

[amotl]

About

GH-10 was finished with a few remaining test cases still (partly) failing. That should be addressed in one way or another.

Details

Revisit those test cases, eventually making them work, and/or assess the impact if corresponding routines do not work correctly, or do not work at all.

Failing / Skipped

• test_delete_restore_experiment_with_runs
• test_set_experiment_tag
• test_insert_large_text_in_dataset_table
• test_get_orderby_clauses

Flaky

• test_log_inputs_with_duplicates_in_single_request. Examples:

  • Run #16710710861
  • Run #6224569395

• test_log_metric_concurrent_logging_succeeds. Examples:

  • Run #6158404123