[report] Allow users to specify commands and files to skip

Adds two new options, `--skip-commands` and `--skip-files`, that allow
users to selectively skip specific command or file collection instead of
having to disable whole plugins to skip those collections.

These options are also exposed via `sos collect`, being gated by a
version of 4.1 since that is the next scheduled release where we can
guarantee this functionality will be present.

Closes: #2203
Resolves: #2271

Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>

8 files changed, 85 insertions, 0 deletions

diff --git a/man/en/sos-collect.1 b/man/en/sos-collect.1
index 7d6f8b87..d4e5e648 100644
--- a/man/en/sos-collect.1
+++ b/man/en/sos-collect.1
@@ -30,6 +30,8 @@ sos collect \- Collect sosreports from multiple (cluster) nodes
     [\-\-password]
     [\-\-password\-per\-node]
     [\-\-preset PRESET]
+    [\-\-skip-commands COMMANDS]
+    [\-\-skip-files FILES]
     [\-s|\-\-sysroot SYSROOT]
     [\-\-ssh\-user SSH_USER]
     [\-\-sos-cmd SOS_CMD]
@@ -261,6 +263,18 @@ defined, or has a version of sos prior to 3.6, this option is ignored for that n
 \fB\-p\fR SSH_PORT, \fB\-\-ssh\-port\fR SSH_PORT
 Specify SSH port for all nodes. Use this if SSH runs on any port other than 22.
 .TP
+\fB\-\-skip-commands\fR COMMANDS
+A comma delimited list of commands to skip execution of, but still allowing the
+rest of the plugin that calls the command to run. This will generally need to
+be some form of UNIX shell-style wildcard matching. For example, using a value
+of \fBhostname\fR will skip only that single command, while using \fBhostname*\fR
+will skip all commands with names that begin with the string "hostname".
+.TP
+\fB\-\-skip-files\fR FILES
+A comma delimited list of files or filepath wildcard matches to skip collection
+of. Values may either be exact filepaths or paths using UNIX shell-style wildcards,
+for example \fB/etc/sos/*\fR.
+.TP
 \fB\-\-ssh\-user\fR SSH_USER
 Specify an SSH user for sos collect to connect to nodes with. Default is root.
 
diff --git a/man/en/sos-report.1 b/man/en/sos-report.1
index 0a571d18..19fb6aae 100644
--- a/man/en/sos-report.1
+++ b/man/en/sos-report.1
@@ -26,6 +26,8 @@ sosreport \- Collect and package diagnostic and support data
           [--log-size]\fR
           [--all-logs]\fR
           [--since YYYYMMDD[HHMMSS]]\fR
+          [--skip-commands commands]\fR
+          [--skip-files files]\fR
           [--allow-system-changes]\fR
           [-z|--compression-type method]\fR
           [--encrypt-key KEY]\fR
@@ -180,6 +182,18 @@ compression-type file extension for example ".zip". ".1", ".gz" etc.).
 This also affects \--all-logs. The date string will be padded with zeros
 if HHMMSS is not specified.
 .TP
+.B \--skip-commands COMMANDS
+A comma delimited list of commands to skip execution of, but still allowing the
+rest of the plugin that calls the command to run. This will generally need to
+be some form of UNIX shell-style wildcard matching. For example, using a value
+of \fBhostname\fR will skip only that single command, while using \fBhostname*\fR
+will skip all commands with names that begin with the string "hostname".
+.TP
+.B \--skip-files FILES
+A comma delimited list of files or filepath wildcard matches to skip collection
+of. Values may either be exact filepaths or paths using UNIX shell-style wildcards,
+for example \fB/etc/sos/*\fR.
+.TP
 .B \--allow-system-changes
 Run commands even if they can change the system (e.g. load kernel modules).
 .TP
diff --git a/sos/collector/__init__.py b/sos/collector/__init__.py
index 3390e0b0..1abb08ae 100644
--- a/sos/collector/__init__.py
+++ b/sos/collector/__init__.py
@@ -84,6 +84,8 @@ class SoSCollector(SoSComponent):
         'preset': '',
         'save_group': '',
         'since': '',
+        'skip_cmds': [],
+        'skip_files': [],
         'skip_plugins': [],
         'sos_opt_line': '',
         'ssh_key': '',
@@ -277,6 +279,12 @@ class SoSCollector(SoSComponent):
                              help=('Escapes archived files older than date. '
                                    'This will also affect --all-logs. '
                                    'Format: YYYYMMDD[HHMMSS]'))
+        sos_grp.add_argument('--skip-commands', default=[], action='extend',
+                             dest='skip_cmds',
+                             help="do not execute these commands")
+        sos_grp.add_argument('--skip-files', default=[], action='extend',
+                             dest='skip_files',
+                             help="do not collect these files")
         sos_grp.add_argument('--verify', action="store_true",
                              help='perform pkg verification during collection')
 
diff --git a/sos/collector/sosnode.py b/sos/collector/sosnode.py
index d0e836a8..75daef22 100644
--- a/sos/collector/sosnode.py
+++ b/sos/collector/sosnode.py
@@ -651,6 +651,16 @@ class SosNode():
         if self.check_sos_version('4.0'):
             self.sos_bin = 'sos report'
 
+        if self.check_sos_version('4.1'):
+            if self.opts.skip_cmds:
+                sos_opts.append(
+                    '--skip-commands=%s' % (quote(self.opts.skip_cmds))
+                )
+            if self.opts.skip_files:
+                sos_opts.append(
+                    '--skip-files=%s' % (quote(self.opts.skip_files))
+                )
+
         sos_cmd = sos_cmd.replace(
             'sosreport',
             os.path.join(self.host.sos_bin_path, self.sos_bin)
diff --git a/sos/report/__init__.py b/sos/report/__init__.py
index a938f706..7838d60a 100644
--- a/sos/report/__init__.py
+++ b/sos/report/__init__.py
@@ -94,6 +94,8 @@ class SoSReport(SoSComponent):
         'list_profiles': False,
         'log_size': 25,
         'map_file': '/etc/sos/cleaner/default_mapping',
+        'skip_cmds': [],
+        'skip_files': [],
         'skip_plugins': [],
         'noreport': False,
         'no_env_vars': False,
@@ -258,6 +260,12 @@ class SoSReport(SoSComponent):
                                 dest="profiles", type=str, default=[],
                                 help="enable plugins used by the given "
                                      "profiles")
+        report_grp.add_argument('--skip-commands', default=[], action='extend',
+                                dest='skip_cmds',
+                                help="do not execute these commands")
+        report_grp.add_argument('--skip-files', default=[], action='extend',
+                                dest='skip_files',
+                                help="do not collect these files")
         report_grp.add_argument("--verify", action="store_true",
                                 dest="verify", default=False,
                                 help="perform data verification during "
diff --git a/sos/report/plugins/__init__.py b/sos/report/plugins/__init__.py
index 4e1a263d..ed0a07ef 100644
--- a/sos/report/plugins/__init__.py
+++ b/sos/report/plugins/__init__.py
@@ -495,6 +495,8 @@ class Plugin(object):
         self.policy = commons['policy']
         self.devices = commons['devices']
         self.manifest = None
+        self.skip_files = commons['cmdlineopts'].skip_files
+        self.skip_cmds = commons['cmdlineopts'].skip_cmds
 
         self.soslog = self.commons['soslog'] if 'soslog' in self.commons \
             else logging.getLogger('sos')
@@ -1080,6 +1082,20 @@ class Plugin(object):
     def _is_forbidden_path(self, path):
         return _path_in_path_list(path, self.forbidden_paths)
 
+    def _is_skipped_path(self, path):
+        """Check if the given path matches a user-provided specification to
+        ignore collection of via the ``--skip-files`` option
+
+        :param path:    The filepath being collected
+        :type path: ``str``
+
+        :returns: ``True`` if file should be skipped, else ``False``
+        """
+        for _skip_path in self.skip_files:
+            if fnmatch.fnmatch(path, _skip_path):
+                return True
+        return False
+
     def _copy_node(self, path, st):
         dev_maj = os.major(st.st_rdev)
         dev_min = os.minor(st.st_rdev)
@@ -1425,6 +1441,9 @@ class Plugin(object):
                 if self._is_forbidden_path(_file):
                     self._log_debug("skipping forbidden path '%s'" % _file)
                     continue
+                if self._is_skipped_path(_file):
+                    self._log_debug("skipping excluded path '%s'" % _file)
+                    continue
                 if limit_reached:
                     self._log_info("skipping '%s' over size limit" % _file)
                     continue
@@ -1577,6 +1596,14 @@ class Plugin(object):
         pred = kwargs.pop('pred') if 'pred' in kwargs else None
         soscmd = SoSCommand(**kwargs)
         self._log_debug("packed command: " + soscmd.__str__())
+        for _skip_cmd in self.skip_cmds:
+            # This probably seems weird to be doing filename matching on the
+            # commands, however we want to remain consistent with our regex
+            # matching with file paths, which sysadmins are almost guaranteed
+            # to assume will use shell-style unix matching
+            if fnmatch.fnmatch(soscmd.cmd, _skip_cmd):
+                self._log_debug("skipping excluded command '%s'" % soscmd.cmd)
+                return
         if self.test_predicate(cmd=True, pred=pred):
             self.collect_cmds.append(soscmd)
             self._log_info("added cmd output '%s'" % soscmd.cmd)
diff --git a/tests/option_tests.py b/tests/option_tests.py
index 56480a58..44e2079f 100644
--- a/tests/option_tests.py
+++ b/tests/option_tests.py
@@ -16,6 +16,8 @@ class MockOptions(object):
     dry_run = False
     log_size = 25
     allow_system_changes = False
+    skip_cmds = []
+    skip_files = []
 
 
 class GlobalOptionTest(unittest.TestCase):
diff --git a/tests/plugin_tests.py b/tests/plugin_tests.py
index 9290003a..0acf07f4 100644
--- a/tests/plugin_tests.py
+++ b/tests/plugin_tests.py
@@ -114,6 +114,8 @@ class MockOptions(object):
     log_size = 25
     allow_system_changes = False
     no_postproc = False
+    skip_files = []
+    skip_cmds = []
 
 
 class PluginToolTests(unittest.TestCase):