ctdb: Leave CTDB log levels at their original defaults

There is too much information with the 'log level' explicitily set to
DEBUG for ctdbd flooding the log file destination. Similarly 'script
log level' option might also spit out more when set to DEBUG. Instead
of removing the options altogether we change the sambacc defaults from
DEBUG to NOTCE and ERROR respectively for 'log level' and 'script log
level' options as specified in man ctdb.conf(5).

ref: https://ctdb.samba.org/manpages/ctdb.conf.5.html

Signed-off-by: Anoop C S <[email protected]>

sambacc/config.py

@@ -312,8 +312,8 @@ def ctdb_config(self) -> dict[str, str]:
         ctdb.setdefault("cluster_meta_uri", CLUSTER_META_JSON)
         ctdb.setdefault("nodes_path", CTDB_NODES_PATH)
         ctdb.setdefault("recovery_lock", CTDB_RECLOCK)
-        ctdb.setdefault("log_level", "DEBUG")
-        ctdb.setdefault("script_log_level", "DEBUG")
+        ctdb.setdefault("log_level", "NOTICE")
+        ctdb.setdefault("script_log_level", "ERROR")
         ctdb.setdefault("realtime_scheduling", "false")
         # this whole thing really needs to be turned into a real object type
         ctdb.setdefault("public_addresses", [])

tests/test_ctdb.py

@@ -438,7 +438,8 @@ def test_ensure_ctdb_conf(tmpdir):
     ctdb.ensure_ctdb_conf(iconfig=cfg.get("ctdb1"), path=path)
     with open(path, "r") as fh:
         data = fh.read()
-    assert "DEBUG" in data
+    assert "NOTICE" in data
+    assert "ERROR" in data
     assert "/var/lib/ctdb/shared/RECOVERY" in data