Problem: A dirty test in test_utils is contaminating the test env

Solution: Added a work around to allow testing of `upsert_validator show`
Added TODOs to detail what is wrong with the broken test
Raising it as a GitHub issue to fix the bad test

bigchaindb/commands/bigchaindb.py

@@ -212,7 +212,12 @@ def run_upsert_validator_show(args, bigchain):
     node_id = new_validator['node_id']
     status = election.get_status(bigchain)
 
-    logger.info(f'public_key={public_key}\npower={power}\nnode_id={node_id}\nstatus={status}')
+    response = f'public_key={public_key}\npower={power}\nnode_id={node_id}\nstatus={status}'
+
+    logger.info(response)
+
+    # TODO: Remove this once tests/commands/test_utils.test_configure_bigchaindb_logging.py is fixed
+    return response
 
 
 def _run_init():

tests/commands/test_utils.py

@@ -72,6 +72,11 @@ def test_configure_bigchaindb_configures_bigchaindb():
      logging.CRITICAL)
 )))
 def test_configure_bigchaindb_logging(log_level):
+    # TODO: See following comment:
+    # This is a dirty test. If a test *preceding* this test makes use of the logger, and then another test *after* this
+    # test also makes use of the logger, somehow we get logger.disabled == True, and the later test fails. We need to
+    # engineer this somehow to leave the test env in the same state as it finds it.
+
     from bigchaindb.commands.utils import configure_bigchaindb
 
     @configure_bigchaindb

tests/upsert_validator/test_validator_election.py

@@ -1,7 +1,6 @@
 # Copyright BigchainDB GmbH and BigchainDB contributors
 # SPDX-License-Identifier: (Apache-2.0 AND CC-BY-4.0)
 # Code is Apache-2.0 and docs are CC-BY-4.0
-import logging
 from argparse import Namespace
 
 import pytest
@@ -144,8 +143,7 @@ def test_get_status_inconclusive(b, inconclusive_election, new_validator):
     assert resp == status
 
 
-@pytest.mark.dev
-def test_upsert_validator_show(caplog, ongoing_election, b, priv_validator_path, user_sk, monkeypatch):
+def test_upsert_validator_show(caplog, ongoing_election, b):
     from bigchaindb.commands.bigchaindb import run_upsert_validator_show
 
     election_id = ongoing_election.id
@@ -157,7 +155,13 @@ def test_upsert_validator_show(caplog, ongoing_election, b, priv_validator_path,
     show_args = Namespace(action='show',
                           election_id=election_id)
 
-    with caplog.at_level(logging.INFO):
-        run_upsert_validator_show(show_args, b)
-        msg = caplog.records.pop().msg
-        assert msg == f'public_key={public_key}\npower={power}\nnode_id={node_id}\nstatus={status}'
+    # TODO: Turn this back on once tests/commands/test_utils.test_configure_bigchaindb_logging.py is fixed
+    # with caplog.at_level(logging.INFO):
+    #     run_upsert_validator_show(show_args, b)
+    #     msg = caplog.records.pop().msg
+    #     assert msg == f'public_key={public_key}\npower={power}\nnode_id={node_id}\nstatus={status}'
+
+    # TODO: Remove everything after this once tests/commands/test_utils.test_configure_bigchaindb_logging.py is fixed
+    msg = run_upsert_validator_show(show_args, b)
+
+    assert msg == f'public_key={public_key}\npower={power}\nnode_id={node_id}\nstatus={status}'