clean: add a new option to clean generated config files

cloud-init generates several config files today that are not cleaned as a part
of `cloud-init clean` operation. For example, cloud-init uses
`/etc/ssh/sshd_config.d/50-cloud-init.conf` to configure ssh daemon which is then
left uncleaned. This change adds a new option `cloud-init clean --configs` that
would clean up these config files as well. Additionally, the cli documentation
has also been updated to document this new option.

Reference:
#1618
Fixes: GH-3240

Signed-off-by: Ani Sinha <anisinha@redhat.com>

cloudinit/cmd/clean.py

@@ -25,6 +25,12 @@
 )
 
 ETC_MACHINE_ID = "/etc/machine-id"
+GEN_CONFIG_FILES = [
+    "/etc/ssh/sshd_config.d/50-cloud-init.conf",
+    "/etc/NetworkManager/conf.d/99-cloud-init.conf",
+    "/etc/NetworkManager/conf.d/30-cloud-init-ip6-addr-gen-mode.conf",
+    "/etc/network/interfaces.d/50-cloud-init.cfg",
+]
 
 
 def get_parser(parser=None):
@@ -77,24 +83,36 @@ def get_parser(parser=None):
         dest="remove_seed",
         help="Remove cloud-init seed directory /var/lib/cloud/seed.",
     )
+    parser.add_argument(
+        "-c",
+        "--configs",
+        action="store_true",
+        default=False,
+        dest="remove_config",
+        help="Remove cloud-init generated config files.",
+    )
     return parser
 
 
-def remove_artifacts(remove_logs, remove_seed=False):
+def remove_artifacts(remove_logs, remove_seed=False, remove_config=False):
     """Helper which removes artifacts dir and optionally log files.
 
     @param: remove_logs: Boolean. Set True to delete the cloud_dir path. False
         preserves them.
     @param: remove_seed: Boolean. Set True to also delete seed subdir in
         paths.cloud_dir.
+    @param: remove_config: Boolean. Set True to also delete cloud-init
+        generated config file. False keeps them.
     @returns: 0 on success, 1 otherwise.
     """
     init = Init(ds_deps=[])
     init.read_cfg()
     if remove_logs:
         for log_file in get_config_logfiles(init.cfg):
             del_file(log_file)
-
+    if remove_config:
+        for conf in GEN_CONFIG_FILES:
+            del_file(conf)
     if not os.path.isdir(init.paths.cloud_dir):
         return 0  # Artifacts dir already cleaned
     seed_path = os.path.join(init.paths.cloud_dir, "seed")
@@ -121,7 +139,9 @@ def remove_artifacts(remove_logs, remove_seed=False):
 
 def handle_clean_args(name, args):
     """Handle calls to 'cloud-init clean' as a subcommand."""
-    exit_code = remove_artifacts(args.remove_logs, args.remove_seed)
+    exit_code = remove_artifacts(
+        args.remove_logs, args.remove_seed, args.remove_config
+    )
     if args.machine_id:
         if uses_systemd():
             # Systemd v237 and later will create a new machine-id on next boot

doc/rtd/reference/cli.rst

@@ -66,9 +66,9 @@ Possible subcommands include:
 :command:`clean`
 ================
 
-Remove ``cloud-init`` artifacts from :file:`/var/lib/cloud` to simulate a clean
-instance. On reboot, ``cloud-init`` will re-run all stages as it did on
-first boot.
+Remove ``cloud-init`` artifacts from :file:`/var/lib/cloud` and config files
+(best effort) to simulate a clean instance. On reboot, ``cloud-init`` will
+re-run all stages as it did on first boot.
 
 * :command:`--logs`: Optionally remove all ``cloud-init`` log files in
   :file:`/var/log/`.
@@ -78,6 +78,8 @@ first boot.
   remove the file. Best practice when cloning a golden image, to ensure the
   next boot of that image auto-generates a unique machine ID.
   `More details on machine-id`_.
+* :command:`--configs`: Optionally remove all ``cloud-init`` generated config
+  files.
 
 .. _cli_collect_logs:
 

tests/unittests/cmd/test_clean.py

@@ -12,7 +12,8 @@
 
 MyPaths = namedtuple("MyPaths", "cloud_dir")
 CleanPaths = namedtuple(
-    "CleanPaths", ["tmpdir", "cloud_dir", "clean_dir", "log", "output_log"]
+    "CleanPaths",
+    ["tmpdir", "cloud_dir", "clean_dir", "log", "output_log"],
 )
 
 
@@ -69,6 +70,33 @@ def test_remove_artifacts_removes_logs(self, clean_paths, init_class):
         ), f"Unexpected file {clean_paths.output_log}"
         assert 0 == retcode
 
+    def test_remove_conf(self, clean_paths, init_class):
+        """remove_config removes config files when True."""
+
+        TEST_GEN_CONFIG_FILES = [
+            clean_paths.tmpdir.join(conf_file)
+            for conf_file in clean.GEN_CONFIG_FILES
+        ]
+        for conf_path in TEST_GEN_CONFIG_FILES:
+            assert conf_path.exists() is False, f"Unexpected {conf_path}"
+            ensure_dir(os.path.dirname(conf_path))
+            conf_path.write(f"#generated by cloud-init\ntouch {conf_path}\n")
+
+        with mock.patch(
+            "cloudinit.cmd.clean.GEN_CONFIG_FILES", TEST_GEN_CONFIG_FILES
+        ):
+            retcode = wrap_and_call(
+                "cloudinit.cmd.clean",
+                {"Init": {"side_effect": init_class}},
+                clean.remove_artifacts,
+                remove_logs=False,
+                remove_config=True,
+            )
+
+        for conf_path in TEST_GEN_CONFIG_FILES:
+            assert conf_path.exists() is False, f"Unexpected file {conf_path}"
+        assert 0 == retcode
+
     @pytest.mark.allow_all_subp
     def test_remove_artifacts_runparts_clean_d(self, clean_paths, init_class):
         """remove_artifacts performs runparts on CLEAN_RUNPARTS_DIR"""
@@ -229,10 +257,14 @@ def fake_subp(cmd, capture):
             return "", ""
 
         myargs = namedtuple(
-            "MyArgs", "remove_logs remove_seed reboot machine_id"
+            "MyArgs", "remove_logs remove_seed remove_config reboot machine_id"
         )
         cmdargs = myargs(
-            remove_logs=False, remove_seed=False, reboot=True, machine_id=False
+            remove_logs=False,
+            remove_seed=False,
+            remove_config=False,
+            reboot=True,
+            machine_id=False,
         )
         retcode = wrap_and_call(
             "cloudinit.cmd.clean",
@@ -264,11 +296,12 @@ def test_handle_clean_args_removed_machine_id(
         """handle_clean_args removes /etc/machine-id when arg is True."""
         uses_systemd.return_value = systemd_val
         myargs = namedtuple(
-            "MyArgs", "remove_logs remove_seed reboot machine_id"
+            "MyArgs", "remove_logs remove_seed remove_config reboot machine_id"
         )
         cmdargs = myargs(
             remove_logs=False,
             remove_seed=False,
+            remove_config=False,
             reboot=False,
             machine_id=machine_id,
         )